Contents

Chapter Purpose Page No.

• Introduce the format of the guide and how to

1. About this Guide 3
use it.

• Explain the purpose of the AI Practice guide,

2. General information introduce key terms and concepts and to 10
inspire people to use AI on their projects.

• Demystify AI
3. An introduction to AI
• Explain the spectrum 15
in project management
• Asking the right questions (prompting).

• Illustrate, mostly by way of example, how the

4. PRINCE2 7 and AI in spectrum of AI can be applied to projects
24
practice through the lens of PRINCE2’s integrated
elements.

• Using PRINCE2’s digital and data

5. Making it happen 39
management approach.

• Explain how individuals and organizations

can prepare for and develop capabilities for
6. Recommendations the pace and direction of change.
43
for practice success • Prompt people to consider upskilling
• Provide recommendations for success based
on the relevant method.

Copyright © 2023 PeopleCert International Ltd.

All rights reserved. No part of this publication may be reproduced or transmitted in any form and by any means (electronic, photocopying, recording or
otherwise) except as permitted in writing by PeopleCert International Ltd. Enquiries for permission to reproduce, transmit or use for any purpose this
material should be directed to the publisher.

Disclaimer
This publication is designed to provide helpful information to the reader. Although every care has been taken by PeopleCert International Ltd in the
preparation of this publication, no representation or warranty (express or implied) is given by PeopleCert International Ltd as publisher with respect as
to the completeness, accuracy, reliability, suitability or availability of the information contained within it and neither shall PeopleCert International Ltd be
responsible or liable for any loss or damage whatsoever (indicatively but not limited to, special, indirect, consequential) arising or resulting of virtue of
information, instructions or advice contained within this publication.

First edition (v1.1 – Sep 2023)

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 2
Chapter 1
About this Guide

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 3
1. About this Guide
1.1 Welcome
Welcome to the first edition of the PRINCE2 7 AI Practice Guide. In the fast-evolving landscape
of artificial intelligence, change is constant. This edition serves as an essential guide for
understanding AI in project management, but rest assured, it will not be the last. Harnessing the
power of AI alongside the PRINCE2 method can be a game-changer, unlocking unprecedented
efficiencies and capabilities in project delivery. Stay tuned for frequent updates that will keep
you at the forefront of global best practices in this exciting and ever-changing field.

1.2 About the Authors

Andy Murray
Andy is the Executive Director of the Major Projects Association, prior to which he spent 30
years as a practitioner, consultant, and assurer of complex projects and programmes for public,
private, and third sector organizations both domestically and internationally. He has a focus on
project governance, the treatment of inherent project complexity, and organizational
development of project capability.

Andy is a respected author, having worked with HM Treasury, Cabinet Office, the Association
for Project Management (APM), and PeopleCert to develop best practice guidance on their
behalf. Publications include PRINCE2, P3M3, Routemap, and the Governance of Co-Owned
Projects. He is the Chair of the APM’s Governance Special Interest Group and is leading on the
Project Data Analytics Task Force.

Andy is a collaborative and inclusive senior leader with a track record of building high-
performing teams both within his organization, across partnerships, and with voluntary groups.

Martin Paver
Martin is the Founder and CEO of Projecting Success, a leading authority on data-driven project
delivery, working closely with professional bodies, government departments, and industry to
transform how projects are delivered. Martin has successfully managed multibillion-dollar
projects, programmes, and portfolios, across industries in the private and public sectors. He
regularly advises governments and the private sector on their transition to data driven project
delivery.

A recognised thought leader, he was the Co-Chair of the Project Data Analytics Task Force and
was named among the most influential people in data by DataIQ100 in 2021 and 2022. Martin
founded the 10,000+ member Project Data Analytics Community and Project:Hack, and is a
member of APM's Data Advisory Group.

Martin is a Fellow of the APM, with a range of qualifications in PRINCE2, Agile, Lean MSP,
portfolio management, and holds an MBA in project management.

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 4
1.3 Format of the guide
Welcome to the PRINCE2® 7 AI Practice Guide. This guide has been developed to help those
working on projects understand how AI can support them with a range of project tasks. It is the
perfect companion to the new AI-ready PRINCE2 7, the latest update to the world-renowned
project management method.

It is split into the following six chapters:

1. About this guide
2. General information
3. Introduction to AI in project management
4. PRINCE2 7 and AI in practice
5. Making it happen
6. Recommendations for practice success

1.4 PRINCE2 7 Summary

PRINCE2, the world's leading project management method, has evolved to meet the needs of
the future. It has been updated to reflect the modern realities of managing successful projects in
the 21st century, incorporating the latest processes, tools, and technologies to stay ahead of the
curve. Best practice has now become even better!
PRINCE2 7 addresses changes in project management and adapts to evolving working
practices and technologies.

There are five integrated elements of PRINCE2 7 which are useful to be reminded of as you
engage with this guide:

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 5
• Principles: the seven principles are the guiding obligations and good practices which
determine whether the project is genuinely being managed using PRINCE2. Unless
these following principles are applied, it is not a PRINCE2 project:
1. ensure continued business justification
2. learn from experience
3. define roles, responsibilities, and relationships
4. manage by stages
5. manage by exception
6. focus on products
7. tailor to suit the project.

• People: an understanding of the needs, capabilities, and motivations of the people

involved and the relationships between them is crucial to how the project is established
and managed. People are at the centre of the PRINCE2 method.

• PRINCE2 practices: these seven practices describe essential aspects of project

management that must be applied consistently throughout the project lifecycle:
1. business case
2. organizing
3. plans
4. quality
5. risk
6. issues
7. progress.

• PRINCE2 processes: these seven processes in PRINCE2 provide the set of activities
required to direct, manage, and deliver a project successfully:
1. starting up a project
2. directing a project
3. initiating a project
4. controlling a stage
5. managing product delivery
6. managing a stage boundary
7. closing a project.

• The project context: PRINCE2 has been designed so it can work within any context,
including in organizational and commercial contexts, delivery method, sustainability
requirements, and project scale. The principles, practices, and processes are applied by
the people involved to ensure that the method is fit for the project context.

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 6
1.5 Example Scenarios
There are four example scenarios used throughout this guide, which demonstrate how AI can
be applied to a wide range of projects using the PRINCE2 method. A short summary of each
scenario is provided below. The following icons represent each of the scenarios that appear
throughout the guide.

Scenario 1: Data Knowledge

Data Knowledge is a company with the mission of leveraging the power of big data
through advanced data analytics solutions. By analysing the wants, needs, desires,
and frustrations of social media users, Data Knowledge can transform raw and unstructured
data into reliable predictions that enable organizations to utilize consumer behaviours and
insights. However, they have not been as effective in using data to their own advantage now
that the company has over 300 staff. In fact, the head of the data science division raised
concerns in the last senior leadership team meeting.

The concerns were regarding the lack of visibility of who is working on what projects, poor
capacity planning, and claims of an overworked team with low morale due to the constant
juggling of commitments and firefighting issues. As a way of addressing this, the data science
division has been authorized to start a project to develop a time-recording solution and a set of
reports on project and task assignments along with resource utilization and capacity planning.
The reports are expected to offer interactive and real-time insights that can enable data-driven
decisions around the utilization and planning of resources.
Data Knowledge has a defined framework for the management of projects. This is expected to
be a simple business change project, taking no more than three months in the current financial
year to complete.

Since Data Knowledge has in-house data analysis and app development capabilities as well as
clear requirements for this project, the project will be led by the head of data science. They will
work with the development division in an iterative and incremental approach. This is to ensure
that the solution can be deployed quickly to several new projects and enhanced during its roll-
out to the rest of the business.

Scenario 2: Louistown City Council

Louistown is a historical town which is home to over 40,000 residents and attracts over
20,000 visitors per year. The city council has been criticized for the lack of investment
and modernization of the town centre. They have recently decided to improve the vibrancy of
the area and offer increased opportunities for business by approving the funds for the
‘Louistown is Open for Business’ (LOB) portfolio.

This is an ambitious portfolio of projects covering the next five-year plan cycle. A new shopping
centre, provisionally named ‘LouisShopping’, is seen as a key project within this portfolio. This is
a complex infrastructure project due to:
• the location of the building and historical nature of the surrounding buildings
• the requirement for the design to be consistent with the environment while offering
modern facilities supported by the latest technology
• the need to include underground parking space
• the importance of minimizing disruption to residents and businesses during the
construction works
• the recent award of ‘green town’ status to Louistown, and the sustainability requirements
this will impose on the development.

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 7
Due to the nature of the work, in which just the construction element is likely to span over two
financial years, the council has approved the strategic business case, the initial high-level
project plan and has appointed BuildyBrick as the main contractor. They have established a
requirement for a traditional delivery method in line with their standard approach to projects.
The general view is that this project is long overdue, and the council would like ‘LouisShopping’
to be open to the public by the end of financial year three.

Scenario 3: Findef
Findef was formed eight years ago with the promise to revolutionize how businesses
identify fraudsters, to protect the reputation of businesses, and to minimize financial
risks. The company experienced substantial growth in the first few years, transforming
from a niche small-medium-enterprise (SME) to one of the leading fintech firms in the field of
financial fraud, working with the world’s largest banks and merchants.

In the last 18 months, Findef’s growth has stagnated. This, coupled with the arrival of new
competitors, is creating discomfort among some investors. In response, the leadership board
recently approved the development of a suite of three innovative cybersecurity products that will
transform the market in financial risk management. Additionally, the leadership board agreed
that Findef will change its commercial and operating models to exploit the new products to their
full potential. The company will now deal directly with the end user rather than through banks
and merchants.

The FindefTwo programme has been established to oversee both the business transformation
and the development of the new products. As the programme requires fresh investment, the
board has established an investment committee that funds and oversees the programme.
The project for the product development will use a hybrid approach of linear-sequential for the
overarching design and deployment of the product set and iterative-incremental for the
development of each product. The project also includes upskilling Findef’s DevOps team in a
new development platform.

To mitigate some risks, the project includes procuring advisors and a marketing agency that
could validate assumptions and share their product experience with the company.

Scenario 4: NowByou
NowByou is a not-for-profit organization that works with local communities to eliminate
discrimination of any kind and help advance the international human rights system
through local and targeted interventions.

NowByou forms, chairs, and equips networking and support groups, where anyone who has
experienced discrimination can share their experiences and thoughts in a safe and empowering
space. Additionally, NowByou is one of the few non-governmental organizations (NGOs) in the
country that works directly with police forces to co-develop campaigns that encourage the
reporting of discriminatory incidents to authorities.

Recent research has shown that discrimination against marginalized groups, particularly the
homeless and refugees, is growing at a concerning rate. The increase in discrimination has led
NowByou to believe that a new six-month campaign is required to address this issue. They have
decided to run the campaign as a project, aiming to both raise awareness of this type of
discrimination and reinforce their own brand recognition (as NowByou is highly dependent on
the goodwill of donors and financial aid from partners).

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 8
The project management maturity of the organization is very low, with no standardized
processes or templates in place. For that reason, NowByou has contacted a professional project
manager with a track record of successful delivery in the not-for-profit sector who will be
responsible for:
• developing the initial project management documentation
• advising and coaching the NowByou director of campaigns, who is sponsoring the
project
• mentoring the staff member who will be appointed as project manager
• advise the project team how to use a variety of media (tv, press, and social networks) for
this, and other future projects.

1.6 How to use the guide

This chapter provides a quick reminder of the five integrated elements of PRINCE2 and
introduces the four scenarios that will be used to illustrate how AI can be applied on projects
using PRINCE2.

Chapter 2 introduces key terms and concepts that will be used throughout this guide and
highlights factors for success.

Chapter 3 explains how the AI term is used loosely by some to mean a range of technology that
assists people in their work and life or by others it is a term used very specifically to mean an
autonomous self-learning supercomputer capable of independent thought. This guide uses the
term to cover this whole spectrum and chapter 3 explains different types of AI from automation
at one end of the spectrum to generative AI at the other end.

Chapter 4 then looks at each of the integrated elements of PRINCE2 to show, mostly by way of
example, how AI can be used to support them. It provides additional considerations on the
application of AI for each of the integrated elements.

Chapter 5 provides a deeper look at a new management product in PRINCE2, namely the
digital and data management approach, and explains how it can be used to consider and then
establish the use of AI on a project.

Chapter 6 recommends how AI can be successfully used on a PRINCE2 project. It considers

the pace of exponential change we are seeing in the development of AI and considers its future
impact on projects, project management, and the profession.

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 9
Chapter 2
General information

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 10
2. General information
2.1 Purpose

Key message
Data analytics and the broader use of digital technologies and AI will transform
how we manage projects. PRINCE2 7 is ‘AI ready’, enabling its use through a
requirement for a digital and data management approach for the project. The
digital and data management approach is integrated in the method through the
PRINCE2 practices and processes.

The purpose of this guide is to provide practical guidance on how AI can be used to support
project management and specifically projects using PRINCE2.

Project management is the application of methods, tools, techniques, and competencies to

enable the project to meet its objectives. Project management involves the planning, delegating,
monitoring, and control of all aspects of the project and the motivation of those involved to
achieve the project objectives within expected performance targets.

Traditionally computers performed programmed tasks and their capability depended on the
directions in their programme. A change of context or requirement would typically require a
change of programme. AI now makes it possible for machines to learn from experience, adjust
to new inputs, and perform human-like tasks. Using deep learning and natural language
processing technologies, computers can be trained to accomplish a wider range of tasks. They
learn by processing large amounts of data and recognizing patterns in the data.

The purpose of using AI in project management is to help humans with project management
tasks, whether that is supporting decision-making through additional insights or automating
routine transactions. In both cases, AI is enabling humans to do things better; better decisions
or more efficient and effective transactions. Ultimately, the use of AI helps reduce or manage
the uncertainty and ambiguity typically associated with projects, which is the very purpose of
project management.

2.2 Terms and concepts

AI is a broad term used to cover a wide range of digital technologies, data analytics, and
machine learning applications.
This guide uses the following terms to describe the AI spectrum:

Automation
The technology-enabled automation of business processes or routine tasks. Some automation
is performed by software robots, or ‘bots’, and is often referred to as robotic process automation
or RPA. RPA is often built into standard common IT platforms or apps, MS Office, or MS Excel.

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 11
Data analytics
The systematic computational analysis of data or statistics. There are four types of data analytics:
• Descriptive, which answers the question, “what happened?” or “what is happening?”
• Diagnostic, which answers the question, “why did this happen?” or “why is it
happening?”
• Prescriptive, which answers the question, “what should we do next?”
• Predictive, which answers the question, “what is likely to happen in the future?”

Generative AI
Generative AI involves the use of deep learning algorithms that can recognize, summarize,
translate, predict, and generate content (text, images, and other media) using very large
datasets. They are often referred to as Large Language Models (LLMs).

Digital technology
Digital technology is not part of the AI spectrum but is used in this guide to describe the use of
technology that complements the use of AI. For example, the use of augmented reality headsets
to assist with stakeholder engagement or design.

These terms and concepts are described more fully in chapter 3.

2.3 Scope
The scope of using AI on projects is limitless, subject to ethical codes applied by the
organization commissioning the project and any governments or non-governmental bodies that
have jurisdiction over the organization.
The use of AI will fall into one of two categories:
• supporting project management, for instance planning
• supporting the delivery of the specialist work, for instance construction.
PRINCE2 is a universal method that can be applied to any type of project. Therefore, it does not
prescribe how to deliver specialist work. Similarly, the primary focus of this guide is how AI can
be applied to project management rather than specialist work, except by way of example using
the scenarios introduced in Chapter 1.

2.4 Success factors

The successful use of automation, data analytics, AI, and digital technology on projects include
the following three factors:
• One: ensure that the use of AI is considered from the outset through developing and
maintaining a digital and data management approach for the project.
• Two: understand the digital and data management capability your project needs through
its lifecycle, its current capability, and possible gaps. Have a plan to close the capability
gap in time for when you need it.
• Three: check the digital and data management approach is consistent with the seven
PRINCE2 principles.

2.4.1 Digital and data management approach

Ensure that the use of AI is considered from the outset through developing and maintaining a
digital and data management approach for the project
The needs, opportunities, and approach for using AI on a project will vary by project. PRINCE2
addresses this by recommending all projects to have a digital and data management approach
specific to the project context. This management approach is developed early in the project

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 12
lifecycle, in outline during the process of starting a project and in detail during the process of
initiating a project. It will be reviewed and updated at every management stage. It enables a
proactive approach to including the use of AI on the project and ensures it is integrated with the
other management approaches, for example risk management.

See Chapter 5 for further guidance on the digital and data management approach.

2.4.2 Digital and data management capability

Understand the digital and data management capability your project needs through its lifecycle,
its current capability, and possible gaps. Have a plan to close the capability gap in time for when
you need it.

The project does not exist in isolation of the organizations involved in the project, therefore
understanding digital and data management capability needs to be in the context of the
organization commissioning the project (the business) and those involved in the project
ecosystem (suppliers and users). There are numerous maturity models for assessing digital and
data management capability. These generally consider:
• Data governance: the process of managing the data lifecycle including how your
organization complies with relevant legislation and approaches ethical considerations
relating to the use of data.
• Data access: if you need large volumes of training data where will it come from? In-
house, bought-in, or via data pooling approaches such as a data trust?
• Data quality: how suitable is your data for it's intended purpose?
• Digital and data security: how secure are your systems and data?
• Digital and data legacy: what will happen with your data when the project completes or
when systems become obsolete?
• Digital and data literacy and skills: the competency your team or organization in using
AI on projects.
• Data and digital leadership: how ambitious and committed is your leadership to
embracing new ways of working and removing barriers for practitioners.
• Systems and technology: how will the infrastructure required to support the use of AI
on projects be sourced? In-house, bought-in, or via a service?

The approach to deploying automation, data analytics, AI, and digital technology on projects will
vary depending on context. For example, the approach for an organization that uses projects for
business delivery, such as a supplier, will be different to an organization that mostly uses
projects for business change initiatives. Likewise, the approach for an organization that heavily
uses data analytics for business delivery, for example an insurance company, is likely to differ to
those organizations where data analytics is emerging across all its functions and operations as
they will already have a core capability and experience to draw upon.

2.4.3 Alignment with the PRINCE2 principles

Check the digital and data management approach is consistent with the seven PRINCE2
principles.

The use of AI on a PRINCE2 project will be set out in its digital and data management
approach. This may include how diagnostic or predictive analytics are used to support decision-
making at each layer within the project (directing, managing, delivering). It may include
automating some PRINCE2 processes or the techniques in the PRINCE2 practices. In some

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 13
cases, it could even mean replacing some techniques with an AI-enabled alternative. Whatever
the approach proposed, it should be checked for alignment with the PRINCE2 principles.

The seven PRINCE2 principles are:

• ensure continued business justification
• learn from experience
• define roles, responsibilities, and relationships
• manage by stages
• manage by exception
• focus on products
• tailor to suit the project.

If the use of AI is not consistent with any of these principles, then the project is no longer being
managed using PRINCE2.

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 14
Chapter 3
An introduction to AI in
project management

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 15
3. An introduction to AI in project management

3.1 Overview
3.1.1 The pervasiveness of AI in work and life
In the vast timeline of technological innovations, only a select few have transformed our lives,
societies, and economies as powerfully as the printing press, engines (steam, internal
combustion, jet), the internet, and the advent of smartphones. Today, Artificial Intelligence (AI),
and particularly Large Language Models (LLMs) like ChatGPT, join this hall of fame of
transformative technologies.
To truly comprehend the transformative nature of AI, consider this: the Internet brought global
information to our fingertips, and smartphones allowed us to carry that world in our pockets. AI,
particularly LLMs, takes this one step further by processing, understanding, and generating
information on our behalf. This is not just about accessing information, but also making it
actionable, personalized, and predictive.
AI in personal life
AI isn't confined to Silicon Valley labs or sci-fi movies anymore. Our daily lives are rife with its
use:
• Cars: from Tesla's self-driving capabilities to safety features like lane departure warnings
or fatigue detection, AI watches out for us on the road.
• Mobile phones: from predictive text to the use of Siri, Alexa, and Google Assistant, our
AI personal assistants, our mobile phones help us schedule, communicate, and
entertain.
• Smart homes: from learning our preferred temperatures to optimizing our energy usage,
smart thermostats are perfect exemplifications of AI at home.
• Shopping: online platforms use AI to suggest products, optimize searches, and even
detect fraudulent activities.
• Health: wearable tech tracks our vital statistics, predicts potential health risks, and even
suggests dietary adjustments.
AI at work
The office isn't immune to the AI revolution:
• Emails: AI-driven spam filters and priority sorting keep our inboxes manageable.
• Calendars: intelligent scheduling tools find the best meeting times considering the
availabilities of all participants.
• Documents and spreadsheets: tools like Grammarly or Excel's forecasting functions
utilize AI to enhance productivity.
• Automation: repetitive tasks such as payment processing are now handed over to
Robotic Process Automation (RPA) systems.
• Chatbots: provide 24/7 customer service and assistance.

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 16
 • Recruitment: AI-driven platforms screen applicants, match job requirements, and even
schedule interviews.
• Banking: fraud detection algorithms keep company finances secure.
Finding balance with AI
Humans excel in creativity, empathy, and contextual understanding, but some tasks are simply
cumbersome or challenging. That's where AI shines:
• Information processing: AI can sift through vast data troves rapidly, something humans
would find time-consuming.
• Pattern recognition: detecting subtle patterns across large datasets is routine for AI, be
it in predicting weather or stock market movements.
• Avoiding cognitive bias: AI can be trained to make decisions free from human biases,
leading to fairer outcomes in areas like recruitment or loan approvals.
In essence, AI is steadily becoming our collaborative partner, amplifying our strengths, and
compensating for our limitations. As we stand at this tech crossroad, the path forward is one of
synergy between human potential and AI capabilities.
Let us not forget that the generation entering the workforce now has never known a world
without the iPhone. These 'digital natives' navigate tech-integrated lives seamlessly. For them,
AI is not a novel concept, but a foundational aspect of their digital experiences.
3.1.2 Applying AI on projects to support successful outcomes
According to the Harvard Business Review, by 2027, approximately 88 million people worldwide
are expected to be working in project management, and the value of project-oriented economic
activity will have reached $20 trillion1. Yet, the rate of successful project delivery in terms of
effectiveness and efficiency remains alarmingly low, at around 35%. This indicates an
extravagant amount of time, money, and opportunity is wasted1,2.
Analysis by Bent Flyvbjerg and Alexander Budzier from Oxford Global Projects on ~12,000
global major projects concluded that the probability of a project being delivered on budget, on
time, and delivering the full benefits defined in the business case is only 0.5%2. That is only 1 in
200 projects delivering on all their promises, so the scope to improve the initiation and delivery
of projects is huge, as is the scale of the economic impact.
How can project data analytics help improve project performance? If a project was a
manufacturing process it could have hundreds, if not thousands, of sensors generating
performance data to be analysed in near real-time with the ability to spot trends, model options,
optimize controls, and maximize yield. However, the typical project dashboard comprises
around 10-20 data items, reported a week or two after month end! Indeed, a RICS report
claimed that 95% of project data is not even used2.

1
https://hbr.org/2021/11/the-project-economy-has-arrived
2
https://majorprojects.org/blog/a-manifesto-for-data-driven-projects/#_ftn1

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 17
A challenge is that project data is often siloed; sat in different organizations, or different
functions are in different departments within organizations. Projects are temporary, so data
(beyond lessons learned reports) is not always preserved after the project closes. This is partly
why the world of project management has been slow to embrace AI and data analytics in the
same way as other fields with shorter cycle times and more repetitive tasks, such as
manufacturing.
This is changing, however. Project professionals are making the shift from projects being
managed as temporary initiatives operating outside of business-as-usual structures (resulting in
slow and manual collection and manipulation of data) to treating them as operational process
which routinely collects and uses performance data. The data is there, it is just that projects and
organizations need to establish the mechanisms and habits to collect and analyse more of it,
more often, and from the start.
Data analytics can help with project level decisions such as understanding changes in
stakeholder sentiment, assessing the productivity of individuals and teams, and understanding
supply chain risks. It can also help with portfolio level decisions, such as deciding which projects
need the most leadership support. Data analytics can do this by understanding which projects,
or which tasks within the project, are predisposed to variance or failure or how many projects
are affected by socio-economic or environmental trends or risks. Drawing parallels with the use
of AI in personal life, imagine having the project equivalent to the lane departure warning
system in our cars: constantly looking at your project, prompting you to take corrective action
ahead of when it is needed rather than after the next checkpoint or stage-end.
Pattern recognition capability in machine learning algorithms enable complex insights to be
uncovered which will improve decision making. This is possible if the algorithms have access to
large amounts of training data through data pooling. Project data permeates all aspects of
project management, so there are opportunities for project data analytics to cover the full
spectrum of project activities. The opportunities are endless.

3.2 The AI spectrum

AI is loosely used to describe a spectrum of computer aided data processing, decision support,
and autonomous decision-making. In this guide we cover this spectrum in four groups, shown in
Figure 1. The AI Spectrum.

AI Spectrum

Data Analytics Generative AI /

Automation
& Visualisation LLMs

Digital Technology
(e.g. sensors, augmented reality, virtual reality)

Figure 1. The AI Spectrum

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 18
Although generative AI receives a significant amount of media attention, it is only one part of the
AI applicable to the project management toolset. For many organizations and project teams,
their immediate opportunities may come from productivity improvements using automation,
improved decision making from insights gained using data analytics or the application of digital
technology specific to their project work, for instance the use of drones as part of quality
assurance.
We explore the AI spectrum in the following four sections.
3.2.1 Automation
Automation provides us with the capability to remove some of the manual, repetitive work. The
range of capabilities are vast, spanning a simple Excel macro through to a complex bot that
mimics human actions by interacting with applications and systems, performing tasks like data
entry, data extraction, and data manipulation, thereby increasing efficiency, and reducing
human intervention.
Over the last 5 years robotic process automation has become commonplace. Operating as a
standalone toolset or integrated into the Microsoft stack via capabilities such as Power
Automate. These tools provide low code, drag-and-drop solutions to automate repetitive tasks,
integrate data and processes across different applications, and streamline business processes
without the need for extensive coding or development expertise.
These capabilities can be nested, enabling workflows to interact with each other and cross
validate. When set up effectively they can help to drive up data quality, consistency, and
reliability.
There is a vast array of free resources available too, via tools such as Python, which can help
with data processing, including removing duplicates, handling missing values, and transforming
data into a standardized format. Python's libraries like BeautifulSoup and Scrapy enable the
extraction of data from websites, allowing users to gather data from various online sources
automatically. Other Python libraries can help with schedule data-related tasks to run at specific
intervals, such as periodic data updates or data validation. Project organizations will evolve from
monthly reporting to near real time data updates on demand. This can also include the
automation of invoices, bill payments, validation of work done, and much more. The
opportunities are vast.
3.2.2 Data Analytics & Visualisation
Project professionals often associate data analytics with reporting, typically via the increasing
use of dashboards and similar visualisations. Take the example of a car. The dashboard
presents the driver with the information relevant to the task. When data exceeds defined
parameters, the driver receives a warning indicator. The service interval is dynamic, based on
how fast the car is being driven. The same will apply to project dashboards. New entrants often
unleash thousands of visualizations that are uncontrolled and fall into disrepute. The more
mature organizations bring governance, role-based insights, with supporting data ecosystems.
But visualizations are only one aspect of data analytics.

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 19
Data analytics is the process of examining, interpreting, and transforming raw data into
meaningful insights and actionable information. It involves using statistical, mathematical, and
computational techniques to discover patterns, trends, and correlations within the data.
Data analytics can be broken down into 4 types:
• Descriptive Analytics: descriptive analytics focuses on summarizing historical data to
provide insights into past events and trends, helping to understand what has happened
and identify patterns in the data. This is where the bulk of the traditional project analytics
is performed, via progress reports, cost reporting, and earned value. However, if we are
reporting on what happened, how can we make effective decisions on what is in the
past? Are we driving the car by looking through the rear-view mirror rather than the
windscreen, combined with a head up display and Google Maps?
• Diagnostic Analytics: diagnostic analytics goes beyond descriptive analytics by delving
into the reasons and causes behind past events, seeking to answer why specific
outcomes occurred based on historical data. If used effectively it can be used to
automate claims and dispute resolution, continual learning, and providing feedback to
future phases.
• Predictive Analytics: predictive analytics uses statistical and machine learning
techniques to forecast future outcomes and trends based on historical data, enabling
organizations to make informed decisions and plan for potential scenarios. This could
range from simple regression through to more complex methods.
• Prescriptive Analytics: prescriptive analytics takes predictive analysis further by
suggesting optimal actions and strategies to achieve desired outcomes, helping
businesses make data-driven decisions by recommending the best course of action.
Using these methods, we acknowledge that the future is not predetermined, and our
intervention can have a significant impact on the outcomes.
Typically, 80% of a data scientist’s role involves collecting, preparing, and cleaning data rather
than the more appealing task of interpreting and extracting insights from data. Depending on the
nature of your project, the reality may be very different, particularly if your project involves
lengthy supply chains. The larger scale, more advanced projects will use common data
environments and integrated data pipelines, but this can be difficult to achieve when
organizations already have their existing tools and infrastructure. There is no one-size-fits-all
solution.
3.2.3 Generative AI / LLMs
AI is a broad field aiming to create machines and systems that can perform tasks that typically
require human intelligence. These tasks can include learning, reasoning, problem-solving,
perception, speech recognition, and decision-making, among others. AI aims to create
machines that can mimic human cognitive abilities and make autonomous decisions based on
data and algorithms. But when do machines become intelligent?

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 20
In 1950, the Turing Test, named after the prominent computer scientist Alan Turing, became the
theoretical method to assess the intelligence of an AI system. In this "imitation game," as Turing
originally proposed, a human participant interacts with both a human and a computer without
knowing their identities. If the computer can convincingly deceive the human participant into
believing it is also human, then it is considered to have passed Turing's test. Recent
developments in LLMs have delivered performance that exceeds that of a human, for specific
use cases.
Machine learning is a subset of AI that enables machines to learn from data without being
explicitly programmed for a particular task. In machine learning, algorithms are used to analyse
and learn patterns from data, and the machine uses these patterns to make predictions, classify
new data, or make decisions. It relies on the availability of data and its quality to improve
performance over time through continuous learning and refinement.
Generative AI is a branch of AI that focuses on creating and generating new data, often in the
form of images, text, audio, or other multimedia content, based on patterns and knowledge
learned from existing data. Unlike traditional AI models that are designed for specific tasks,
generative AI models can produce original content and can be used for tasks such as image
synthesis, text generation, music composition, and more.
In a project delivery context, a good example is generative design in construction. A capability
that leverages computational algorithms and artificial intelligence to automatically generate and
explore multiple design alternatives for a given project definition. It involves defining a set of
design constraints and objectives, such as material properties, budget, safety standards, and
then using algorithms to iteratively create and evaluate numerous design options. The goal is to
find the most efficient, structurally sound, and optimized design that meets the specified criteria,
allowing architects and engineers to explore creative and innovative solutions that might not
have been discovered through traditional manual design processes. These capabilities were
initially based on logic or rules, but more recently they are exploiting generative AI.
If we can already design bridges and motorways using generative design, how long will it be
until we can generate options that balance and optimize for schedule, return on investment,
carbon, safety, cost, and a variety of other factors? The technology exists, but the profession
lacks the integrated data and models to underpin it; but this is changing.
Large Language Models (LLM) are a type of generative AI. LLM is trained on vast datasets,
often containing billions of words, to recognize patterns, grammar rules, and semantic
relationships between words and sentences. This training enables it to predict the most likely
next word in a sentence, generate coherent paragraphs, and understand the context of words in
different circumstances. These capabilities can be enhanced further when combined with other
AI methods, such as Midjourney or Stable Diffusion that use a LLM to understand the text
prompt and then diffusion models to gradually subtract noise and yield a picture that resembles
images from its source data. Users assist in continually improving the quality of the outputs.
Within a project context, LLMs can be trained as chatbots to answer questions around a corpus
of documentation. This could range from PRINCE 2 documentation and templates through to

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 21
HR policies, accounting, and tax. By training the LLM on project data, the chatbot will be able to
answer questions relevant to that dataset, such as comparisons between different reporting
periods, outliers, trends, and much more.
Capabilities such as Microsoft Copilot will take these capabilities further by extracting the results
and automatically visualizing them within a dashboard in PowerBI. These capabilities will
fundamentally change how project professionals extract, integrate, validate, interrogate, and
visualize data.
3.2.4 Digital Technology
The use of digital technology will vary by project use case. Construction projects will
increasingly use cameras, IOT sensors, wearables, and much more to provide real time insights
into performance on the project. Capabilities such as MAFIC help to measure productivity of
each member of the workforce and use this to inform different shift patterns, training, or other
interventions. A company called Buildots uses helmet mounted cameras to collect imagery on
site, integrate with the 3D model, and measure progress against plan, rescheduling in near-real
time. The more successful organizations will be those who can integrate this data end-to-end,
including across their supply chain to optimize, successfully navigate constraints, and respond
to real time challenges.
Transformation projects will make greater use of sensemaking capabilities. Gauging where the
project is delivering results and where interventions are needed. This could include capabilities
such as the MeetingQuality add-on to web conference platforms that provides insights into team
cohesion, surfaces lead indicators of misalignment, and much more.
Augmented reality will help to blend the real and virtual worlds. Assisting everything from
walkthroughs, on-demand documentation, methods, and much more.

3.3 How PRINCE2 is AI ready

PRINCE2 has been designed to be universal and flexible. That means it can be applied
to any type of project, in any type of organization. It also means it can be applied and
tailored to be supported by any type of tool, whether an off-the-shelf project
management tool, an in-house tool, or a collection of tools both off-the-shelf and in-
house. The use of tools to support the directing, managing, and delivery of the project is
captured in each of the management approaches, for example the risk management
approach, and is then considered in a coherent way in the digital and data management
approach.
Each management product in PRINCE2, such as a plan, describes the management
information expected but not how it is obtained, generated, or presented. The emphasis
is on how the information is used to support decision-making. This is reflected in the
PRINCE2 processes, with each process activity showing the information from which
management products is to be used as an input, should be checked, or reviewed, or will
be generated as an output.

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 22
The digital and data management approach establishes how the project will get, use,
and maintain data to enable good decisions that will initiate and deliver the project in a
controlled way. This can include any use of technology across the AI spectrum
introduced in this chapter.
Remember, PRINCE2 describes what needs to be done, not how it is done; whether by
hand, semi-automated, automated, or AI assisted.

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 23
Chapter 4
PRINCE2 7 and AI

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 24
4. PRINCE2 7 and AI
4.1 Overview
This chapter looks at each of the five integrated elements of PRINCE2 to show, mostly by way
of example, how AI can be used to support them. It provides further considerations on the
application of AI for each of the integrated elements, comprising of:
• PRINCE2 principles
• PRINCE2 practices
• PRINCE2 processes
• People
• The project context.
The scenario examples and the considerations for each of the five integrated elements of
PRINCE2 in this chapter help readers to explore how the spectrum of AI (described in chapter
3) can be used on their project.

4.2 PRINCE2 Principles

The PRINCE2 principles are the guiding obligations that determine whether the project is
genuinely being managed using PRINCE2. They ensure effective application and tailoring of
PRINCE2 to any project.
The seven PRINCE2 principles are:
• ensure continued business justification
• learn from experience
• define roles, responsibilities, and relationships
• manage by stages
• manage by exception
• focus on products
• tailor to suit the project.

4.2.1 Overview of where AI can help

The table below shows where AI can best help which principle.
Principles
Data Analytics &
Automation Generative AI / LLMs Digital Technology
Visualisation
Manage by stages Ensure continued Learn from experience Any
business justification
Manage by exception Tailor to suit the
Define roles, project
responsibilities, and
relationships

Focus on products

For decades, the project profession has been capturing lessons learned. Project practitioners
capture lessons in their lessons log and then reflect on those to document in a lesson learned
report of what went well and how to improve. The challenge has always been how to take the
lessons in those reports and turn them into action. In 2009 PRINCE2 introduced the principle of
learning from experience as with a ‘pull’ rather than ‘push’ philosophy. That is, it is the duty of

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 25
everyone working on a project to seek lessons from other similar tasks, work packages or
projects and apply them to their work.
If we now imagine this principle in the context of a Google Map or Waze. We can use AI to help
us navigate a route through our project, to optimize and help us with key decisions at specific
points.

The amount of data being generated and captured by people, projects, and organizations is
growing at an exponential rate. Making sense of huge volumes of data is a challenge for
humans, which is where advanced data analytics and AI can help. The evidence-based insights
gained can help inform the development of the PRINCE2 method and make best practice even
better. The codification of project management insights into global best practice enables a
generic approach to be provided. AI will enable us to turbocharge these assets to be ‘tailored to
suit the project’, which could range from suggesting a logic flow or duration for a scheduled work
package through to scenario-based recommendations on courses of action. Lessons learned is
no longer about the abstracted insights; we can leverage the entirety of our experiences.

We can ingest large volumes of data into LLMs to gain applicable insights. For example, we can
ask the model what the key challenges are likely to be on our next work package. We can use
advanced data analytics to drill into the data to understand probabilities, which mitigations have
previously been most successful, and recommend appropriate tolerances. But the user needs
enough experience to understand when the answer is logical and when the model is
‘hallucinating’, or generating false information. There could be several versions of this model.
One that uses public data and provides the best answer based upon the data it has been
trained on. Or an organization-specific model that is trained on their own corpus of data, which
increases relevance. Imagine that OpenAI is a model that has been trained to speak English
and the organization-specific model is taught the local dialect.

4.2.2 Scenarios
Louistown City Council: The project manager is leading three different projects and
is overstretched. She is time poor. The principle of managing by exception is incredibly
important to her, but she is notified of exceptions too late to positively influence them.
She decides to deploy advanced data analytics and AI to help her with the assurance activities.
The team use data analytics to understand the level of process adherence, such as whether the
team plans are being managed well, by deploying Python scripts on project records.
BuildyBricks has recently bought an AI planning tool which has been trained on hundreds of
thousands of construction projects to predict the schedule and cost profile based on the
breakdown of the work. They have set up automatic alerts in their system if the schedule or cost
is forecast to exceed tolerance thresholds defined by the algorithms. These alerts are dynamic,
adjusting to the stage in the project lifecycle. The Council is also exploring how to integrate the
planning data with other data, such as technical queries or early warning notices, to improve the
performance of the algorithms.

NowByou: Lacks the professional project delivery skills to deliver the project and their
charity donations have been reduced since COVID. They can no longer afford to
recruit external support for this role. They appointed Sue, their campaign manager, to
manage the project. She uses a LLM to guide her step by step through the process and draft
the project initiation documentation, which she refines specifically for her project. She also asks
the model for advice on the most suitable campaigns and measures of effectiveness. She learns
about Google Analytics and similar KPIs, pulling data out via an API into PowerBI to create a
real time dashboard on the effectiveness of the campaigns.

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 26
4.2.3 Considerations
The PRINCE2 principles underpin the method and are reflected in each of the integrated
elements of people, practices, processes, and project context. Therefore, it is more a case of
how AI is applied to the project whilst being consistent with the principles than it is of where AI
helps each of the principles.

4.3 PRINCE2 Practices

Practices are aspects of project management that must be applied consistently and throughout the
project lifecycle. The practices require specific treatment of that aspect of project management for the
PRINCE2 processes to be effective.

The seven PRINCE2 practices are:

• business case
• organising
• plans
• quality
• risk
• issues
• progress.

PRINCE2 practices describe aspects of project management that must be continually addressed as the
project progresses through its lifecycle. PRINCE2 provides guidance for effective management and
suggested techniques for each practice.

4.3.1 Overview of where AI can help

The PRINCE2 practices are designed to help the project management team apply and tailor the method,
principally through a suite of management approaches (such as the benefits management approach).
Each management approach defines the procedures, techniques, and standards to be applied and the
responsibilities for that aspect of project management to be effective. LLMs such as OpenAI’s ChatGPT
can generate the draft content for these management approaches based on information provided in the
project brief. This is not the same as using a template as the drafts generated will be based on the
project’s context. The better trained the LLM is, the better the draft of the management approach will be.

This is why some organizations are using the open-source LLM platforms and training it on sector or
organizational data.

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 27
The table below highlights where the AI can help each of the PRINCE2 practices across the AI spectrum.

Data Analytics & Generative AI / Digital

Automation
Visualisation LLMs Technology

Diagnostic Generate draft Data pipelines for

/predictive business case benefit measures
Business case (modelling of based on project (pre, during, post
options and brief project)
impacts)

Workflow of Blockchain
governance regarding
arrangements, for commercial
Organizing example approval arrangements
authority
Impact on roles

Logic, Generate draft BIM

sequencing, plans based on
Plans
resource project brief and
optimization business case

Assess quality of Generate draft AR/VR quality

Quality outputs vs a product assurance
training data set descriptions

Diagnostics (root- Generate initial

causes and risk register
responses) based on project
brief and
Predictive business case
(modelling of
Risk
‘what if’ Recommend
scenarios) responses to
risks and
Move towards contingency
variance allocation
analytics

Diagnostics (root- Recommend

Issues causes and responses to
responses) issues

Automated, real- Generate Interrogate Sensors for

time reporting forecasts progress via assessing
natural language progress
Generate
Progress checkpoint
reports, highlight
reports, and end
stage reports
using dashboards

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 28
4.3.2 Scenarios

Organizing Practice
Findef: Has a complicated project team structure, with a significant amount of external
support. They have a well-defined RACI, but the decision-making process is regularly
impeded because of problems with interfaces between corporate systems that have
been set up for ‘business as usual’ rather than project work. Findef implemented a decision-
making workflow process based on the project RACI that assigns tasks to specific people, which
in turn trigger other actions when completed. The automation engine performs an assurance
check on each step of the workflow to check for compliance or quality issues, ensuring that the
next step has the best chance of success. The project manager has a real-time dashboard that
tracks the progress of key decisions. The decision-making workflow system highlighted that
Brian from the marketing agency is consistently late approving designs due to being overloaded.
The project manager uses this evidence to renegotiate the level of support provided by the
agency.

Business Case Practice

Data Knowledge: Has outline approval to begin the project, but Emily who would have
written the detailed business case is on sick leave. The data science team ingest 20
other business cases into their localized LLM and asked it to draft a detailed business
case for this project based on the project brief. Within one hour, they have a draft which
they circulate to relevant stakeholders to refine. By the end of the day, they submit the detailed
business case to release the required resources. The approvals team use their own LLM to
identify potential challenges for this project, note that the majority of these have been identified,
and authorize the team to proceed.

Risk Practice
Data Knowledge: The project manager intends to hold a risk workshop to help
prepare the risk management approach and the initial risk register. To help guide the
workshop, the project manager uses ChatGPT to explore likely risks and potential
mitigations.

Prompt: We want to develop a time recording solution for our projects. What risks should
we prepare for given our context?

Response: Given the context provided, Data Knowledge is attempting to internalize its own
expertise by developing a time-recording solution to address concerns raised about resource
management.

Here are some potential risks they might face:

• Unclear Requirements or Scope Creep: even with clear initial requirements, there might
be changes or expansions in the scope as the project progresses. This can extend the
timeline and increase the costs.
• Technical Challenges: implementing a real-time, interactive reporting system can
introduce technical difficulties, especially with data integrations and ensuring data
accuracy.
• User Adoption: not every staff member might be enthusiastic about tracking their time or
using a new tool. Resistance can reduce the effectiveness of the system.
• Data Privacy Concerns: time-recording can be seen as invasive or micromanaging.
There might be concerns about how the data is stored, accessed, and used.

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 29
 • Resource Overallocation: with the data science division handling both their regular tasks
and this project, they might become overburdened, leading to potential delays or
reduced quality in both areas.
• Inadequate Testing: given the short timeframe of three months and an iterative
approach, there’s a risk of inadequately testing the system, leading to potential bugs or
inefficiencies when it’s rolled out.
• Dependency on Key Personnel: if the project is highly dependent on the head of data
science or a few key individuals, any absence or turnover can jeopardize the project
timeline.
• Mismatch of Expectations: the leadership might have high expectations regarding the
system’s capabilities, and if not managed, there could be disappointments with the end
product.
• Incomplete Integration: given that the solution is expected to provide insights on various
aspects (task assignments, resource utilization, etc.), there’s a risk of some elements not
being integrated seamlessly.
• Underestimation of Project Complexity: as this is seen as a “simple business change
project,” there’s a risk of underestimating its complexity, leading to unforeseen
challenges.

Mitigation Strategies:
• Frequent Communication: regular check-ins with stakeholders to ensure everyone is
aligned with the project’s progress and any changes in scope.
• User Training: provide adequate training to staff members to ensure they understand
and can efficiently use the time-recording solution.
• Feedback Loops: given the iterative approach, gather feedback from early users and
adjust accordingly.
• Clear Data Policies: clearly outline and communicate how the data from the time-
recording solution will be used and ensure all measures are GDPR compliant (or
compliant with relevant data protection laws).
• Resource Backup: have a backup plan or additional resources available in case of
unforeseen challenges or overburdened team members.
• Thorough Testing: even in an iterative approach, ensure each iteration is rigorously
tested before deployment.
• Set Clear Expectations: manage expectations by clearly defining the project’s objectives,
scope, and limitations upfront.

The project manager is particularly concerned about the suggested risk of incomplete
integration, so sought further insight from ChatGPT on this area.

Prompt: How can Data Knowledge reduce the risk of incomplete integration if it uses an
agile approach to product development?

Response: Using an Agile approach to product development offers several strategies and
practices that can help Data Knowledge reduce the risk of incomplete integration. Here’s how:
• Incremental Development: Agile promotes the development of small, incremental
changes. By building and integrating features piece by piece, you can ensure each part
works correctly before moving on to the next.

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 30
 • Continuous Integration (CI): CI is the practice of frequently integrating code changes into
a central repository. After changes are integrated, automated tests are run to ensure the
new code integrates correctly with the existing system.
• Regular Demos and Feedback Loops: at the end of each sprint, the team showcases the
implemented features to stakeholders. This provides an opportunity to check if the
integration meets business needs and to gather feedback.
• Cross-functional Teams: Agile teams are often cross-functional, meaning they have
members with different skills and expertise (for example developers, testers, business
analysts). This helps ensure that the integration aspect is considered from various
perspectives.
• Automated Testing: Agile emphasizes automated testing, which can detect integration
issues early. By having a suite of regression tests, the team can be confident that new
code hasn’t broken existing integrations.
• Definition of Done (DoD): the DoD is a clear and concise list of requirements that a
software increment must adhere to for the team to call it complete. Ensuring that
“integration” is a part of this definition will guarantee that features are not considered
“done” until they are fully integrated.
• Frequent Refactoring: Agile promotes regular refactoring to improve code quality. By
continuously improving the codebase, the team can ensure that integrations are kept
clean and efficient.
• Pair Programming: two developers working on the same piece of code can catch
integration issues earlier and can discuss and address potential integration challenges
on the spot.
• Daily Stand-ups: these are short daily meetings where team members discuss what they
worked on the previous day, what they plan to work on today, and any blockers they
might have. Any integration issues or concerns can be quickly highlighted and
addressed.
• Use of Integration Tools: tools like Jenkins, Travis CI, or CircleCI can help in automating
the integration process, running tests, and ensuring that all components work
harmoniously.

The project manager delves into a few more of the suggested risks with follow-up questions.
They use the ChatGPT responses to add further project specific content to the project’s draft
risk management approach and initial risk register. The suggested risks are aligned to Data
Knowledge’s own risk categorization, based on the PESTLE technique as recommended in
PRINCE2. The draft risk management approach and risk register are reviewed and refined in
the risk workshop. Ahead of the risk workshop, the project manager consults with the manager
of the DevOps team to find out if or what integration tools the company is using.

4.3.3 Considerations
When developing the digital and data management approach for the project, consider the
following in relation to AI enabled PRINCE2 practices:
• Specialist help may be needed to establish workflows between organizations on different
corporate systems.
• How does the project (or organizations involved) get access to the volumes of training
data required, particularly for forecasting. Is this available in-house if the organization
has a large portfolio of projects? Or would the project or organization need to make use
of data trust or benchmarking service?

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 31
 • Be aware of whether the project has siloed or connected data. The key insights come
from the connections in project data, for example linking schedule and cost data.
• How will the project ensure that it is compliant with legislation or business policies
relating to Personally Identifiable Information (PII) and commercially sensitive data?

4.4 PRINCE2 Processes

PRINCE2 is a process-based approach for project management. There are seven processes in
PRINCE2 that provide the set of activities required to direct, manage, and deliver a project
successfully.
The seven PRINCE2 processes are:
• starting a project
• directing a project
• initiating a project
• controlling a stage
• managing product delivery
• managing a stage boundary
• closing a project.

4.4.1 Overview of where AI can help

A process in PRINCE2 is a structured set of activities that define the sequence of actions and
their inputs and outputs to achieve a specific objective. Therefore, using automation to support
processes improves their repeatability, consistency, and efficiency.

The table below highlights where the AI can help each of the PRINCE2 processes across the AI
spectrum.

Processes
Data Analytics &
Automation Generative AI / LLMs Digital Technology
Visualisation

Establish workflow to BI dashboards to

automate triggers from show status, diagnose Embedding PRINCE2
one process to the process performance into a project
next issues, and prescribe management system.
improvements.

4.4.2 Scenarios

Controlling a stage Process

Data Knowledge: Has a key challenge relating to capacity planning. They created a
low code time-recording application that allows employees to log their work hours,
project assignments, and tasks. In parallel, they created a list of resources, skills, and
constraints. They deployed the open-source PuLP algorithm in Python to solve optimization
problems, including resource allocation and capacity planning, using linear programming
techniques. This provided a first order analysis of how best to assign resources. They
developed the capability further by leveraging time sheet data to understand the relative
productivity of each member of the team and feed this parameter into the resource model. They
can now dynamically assign resources taking account of emerging priorities and constraints.

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 32
Starting a project Process
Findef: is an organization that is expert in understanding financial risks. A large
component of their in-house capability concerns the deployment of graph databases
and AI to identify patterns that flag the risk of fraud. They decided they wanted to use
their pattern recognition capability on prioritizing the product roadmap in the FindefTwo
programme. They wanted to map the projects, products, resources, constraints, and market
feedback. A multi-dimensional challenge, with an added time component. The team utilized
internal expertise to develop a graph database that mapped the interconnections between each
of these attributes, then added additional richness around key project variables for each new
product such as cost, variance, and more. They were able to query the database and rapidly
understand the composition of their product roadmap, which automatically fed into the
programme plan and provides the trigger for the starting a project process for each new product
to be developed. The trigger from their graph database and AI engine provides a project
mandate containing a draft project product description based on the understanding of desired
features and optimal market timing.

4.4.3 Considerations
When developing the digital and data management approach for the project, consider the
following:
• As an organization aggregates and collates data it becomes a significant risk should it
ever be compromised. Always consider how to develop guardrails and data partitioning
to mitigate this risk.
• When deploying AI, consider how to make best use of patterns in the supporting data. A
project may be entirely unique when looking through a product lens, but there may be a
lot of similarities with other projects when viewing through a functional lens.

4.5 People
In PRINCE2, the ‘people’ element of the method covers those who are working on a project and
the relationships between them, as well as those impacted by a project.
The people element of the method includes guidance on:
• leading successful change
• leading successful teams
• communicating.
The success of a project will depend on how well the project establishes strong relationships
with the organizations that support its delivery, as well as how it fits into the organizational
ecosystem to deliver the change. Projects combine a unique set of people from across the
business, user, and supplier communities for a limited period. To successfully deliver the project,
all levels of leadership must establish a project culture that motivates people.

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 33
4.5.1 Overview of where AI can help
AI can help project teams understand the complexity of such relationships, automate the
collection and dissemination of information which aids communication and insights, or use
augmented reality or virtual reality to simulate the project product or the journey to deliver it (the
project plan) to engage stakeholders as part of the change management approach.

The table below highlights where AI can be applied to PRINCE2 people across the AI spectrum.

People
Data Analytics and Digital technology /
Automation Generative AI / LLMs
visualisation other
Consultation
Consultation Generating the draft
(understanding)
(collation) communications
management
Sentiment analysis of Support for hybrid
On-boarding and off- approach and the
stakeholders across working
boarding change management
the project ecosystem
approach
and organizational AR/VR simulation of
Responding to
ecosystem project product or
enquiries or FAQs Generating
(engagement) project plan
through a project communications
(engagement)
chatbot content
Network analysis (of
relationships)

4.5.2 Scenarios

NowByou: Has planned an initial consultation to inform their main campaign. They
wanted to use Generative AI and RPA to support the consultation work and explained
their use in their communications management approach. Their early market
engagement identified four personas they wanted to target. These ranged from an HR
professional who has a deep understanding of discrimination through to someone who is
unconsciously discriminating against others. They used ChatGPT to develop six scenarios and
then set up storytelling chatbots on their website with which consultees could interact to explore
different types of discrimination. They used a Microsoft form to capture feedback to enable them
to better target their messaging. They then used ChatGPT to refine and hyper-personalize the
messages for the different personas, taking account their level of awareness and
understanding, for the main campaign.

Data Knowledge: Has a problem with its team being overworked, but it does not have
the volume of work in the pipeline that justifies the need to work long hours. Something
is going wrong somewhere. Emily, their project lead, works with the data science team
to extract activities and durations from diaries. She integrates this with the sprint plan and
deliverables. Two weeks later, she concludes that the team is spending over 25% of their time
in meetings. She then used an app to understand what value people were getting from the
meetings and found that attendance had grown over time, and many people were only there
because they were invited. Emily asked the meeting organizers to rationalize attendance to key
decision makers. She also used AI to summarize the key outcomes from each meeting and
circulate via Teams channels to keep everyone appraised of progress. She reduced the 25% of
time in meetings to 10%.

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 34
4.5.3 Considerations
When developing the digital and data management approach for the project, consider the
following in relation to people:
• Privacy of stakeholders in the project ecosystem and organizational ecosystem and
whether anonymisation, pseudonymisation, or similar approaches are required.
• Security of personal data and insights. Role based access will be required.
• Ethical use of insights gained.
• Whether there are any blind spots in insights or engagement due to some stakeholder
groups being less digitally connected than others.
• Understand the diversity of the training data used for LLMs when generating content as
it may inadvertently reduce diversity in its output. For example, some open access
generative AI platforms for artwork have been shown to create images of mostly white
men for manager roles.
• Danger of using data to target specific groups or people.

4.6 The project context

The principles, practices, and processes in the PRINCE2 method are applied by the people
involved to ensure that the method is fit for the project context. PRINCE2 has been designed so
it can work within any context.

There are five common contexts included in the PRINCE2 book to illustrate how the method can
be applied or tailored:
• organizational context, such as sector, capability, culture, policies
• commercial context, such as contracts, relationships
• delivery method, such as lifecycle model, development techniques
• sustainability requirements, such as business drivers and priorities
• project scale, such as complexity, duration.

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 35
4.6.1 Overview of where AI can help

Processes
Data Analytics Digital
Generative AI /
Automation and technology /
LLMs
visualisation other
In-house project
Development of
management
in-house LLM to
system based on
include business
the use of
Organizational policies,
PRINCE2
standards, and
management
methods
products

Automation of
contract notices
(such as changes,
early warning Blockchain for
notices, claims) Drafting of procurement,
Analysis of tender
Commercial procurement fulfilment,
submissions
Automating documentation invoicing, and
invoicing and payments
payments based
on work package
approval

Delivery Data analytics of

method user feedback to
(agile) inform next sprint

Use of sensors on
project
Sustainability
equipment/plant to
requirements
minimize idling or
inefficient use

Small projects will

rely on off-the-
shelf or open-
source tools

Project scale Large projects

may have a
centralised
business providing
data analytics
services to
projects

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 36
4.6.2 Scenarios
Louistown City Council: The government uses a risk potential assessment to
characterize the complexity of a project, which shapes the level of assurance and
support required. Louistown City Council wanted to deploy this methodology to assess
how best to assure their portfolio of projects. But they were concerned that the approach wasn’t
sufficiently dynamic to take account of the emerging conditions such as high rates of inflation,
poor availability of some building materials, the eventuality of another pandemic. They
developed a simple app that automated the risk potential assessment, which identified specific
areas of focus. They also set up an automation script that regularly queried the credit rating and
Companies House data for key suppliers. They also linked this to public data from the county’s
Contracts Finder website to understand how much government work each supplier had won.
This provided a lead indicator of whether the contractor had too much work on (which could
adversely impact resources available to the project) or too little, which could be an indicator of
financial distress. Automatic triggers were set up to initiate additional due diligence for stressed
suppliers.

The Council wanted to secure best value for money for the taxpayer by ensuring they used the
right form of contract for each project and contractor. They realized that a one-size-fits-all
approach to contracting was not appropriate. They developed a simple decision tree-based
algorithm to select the best form of contract for each project. They used automation to extract
key parameters for each prospective contractor from emails, company registration and trade
association websites to populate the contract boilerplates, before performing an automated
assurance check prior to issue. The automation process extracted the key parameters from
each contract into a portfolio management tool, recording details such as project start, finish,
value, key milestones etc. This was integrated into PowerBI and regularly reviewed by the CEO
to understand overall cashflow. The team used Python to extract invoice values from emails,
enabling them to report forecast against actuals. They deployed AI to learn from this data to
provide a dynamic forecasting capability, reflecting performance on the ground.

BuildyBricks is the prime contractor on the Louistown project. They have bought a capability that
uses AI and LIDAR that runs daily scans of each room on the construction site. They take these
images and integrate them with their 3D model. They can see where actual progress has been
made against the planned variance and flag any divergence. This allows them to automatically
replan the following day’s activities, plan logistics, and optimize work phases. They use this data
to forecast end dates and report, in near real time, to the client. As a main contractor,
BuildyBricks appreciate that the key to a long-term successful relationship with the Council is to
drive up productivity, reduce cost, and improve delivery confidence. They are experts in Lean
and the Theory of Constraints. They have a long-term strategic plan to create a fully integrated
model of their construction sites that leverages real time data feeds. They have deployed a hard
hat capability that uses sensors to track head movement to understand personal productivity.
They run experiments to understand the impact of shift patterns and weather on personal
output. They move to a two-break shift pattern model and drive-up workface productivity by
14%. They use machine learning to integrate pit lane deliveries with crane availability and
logistics to ensure optimized ‘flow’ across site. The key to all of this is their level of ambition,
vision, skills, and end-to-end data integration.

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 37
4.6.3 Considerations
When developing the digital and data management approach for the project, consider the
following in relation to project context:
• Consider the vast range of data sources that may be available to you. Some are
available via third party data aggregators. Others are available directly and can be
accessed via APIs or web scraping tools such as BeautifulSoup.
• Do not just automate the existing process. Take a step back to reimagine the process
end-to-end. There may be a better way to deliver the outcomes.
• Build error checking and resilience into the process from the outset.
• Consider maintenance of the code too. Upgrades in libraries and code may stop the
application from working, requiring the code to be reconfigured.

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 38
Chapter 5
Making it happen

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 39
5. Making it happen
5.1 Overview
PRINCE2 7 is ‘AI ready’, enabling its use through the requirement for a digital and data
management approach for the project.

The digital and data management approach is developed in outline in the starting a project
process, completed in the initiating a project process and updated for each stage of the project
through the managing a stage boundary process. The digital and data management approach
underpins PRINCE2’s progress practice in terms of establishing a data-driven approach to
understanding current and forecast progress, but it is also used by each of the other six
practices. It is the management product that brings it all together so that the project has a
considered and coherent approach to the use of AI and technology.

5.2 The digital and data management approach

Purpose
The purpose of the digital and data management approach is to describe:
1. how digital technology will be used to support project management (for example
project controls) and project work (for example the use of digital twins for design
and construction)
2. how data and information will be created, used, and managed across the project
ecosystem and through the project lifecycle and afterwards.
High-level content
Scope: describes what data is to be managed.
Digital technology requirements: an analysis of the digital technology the project will
need by considering the automation and facilitation of project management activities and
any specialist technology required to support the specialist project work. In some cases,
this may involve the use of systems already established by the business and suppliers. In
other cases, different or additional systems may be needed.
Data management requirements: an analysis of the data the project will generate or
needs by considering:
• What data or information is needed or will be created?
• Where will the data come from?
• Where will the data be held?
• How will the data be kept secure?
• What level of privacy is required?
• How the data or information will be analysed, curated, and presented?
• Who will have access or rights? How integrity will be ensured?
• How will data quality be assured?
• What will happen to the data after the project is completed?
Digital and data management procedures: description of any digital and data
management procedures to be used on the project (based on the analysis of the digital
and data management requirements).

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 40
 Responsibilities: who will be responsible for the roles associated with the digital data
management approach.

Timing formal digital and data management activities: for example, the
implementation of systems, the timing of any data audits, or the provision of a data room
to support procurement activities.
Supporting tools or systems: describes what tools or systems will be used for digital and
data management.
Standards: describes any digital and data management standards required for the
project.
References: for any associated documents or products.

5.3 Considerations
5.3.1 Data management
The product outline for the digital and data management approach includes some
considerations relating to data:
• What data or information is needed or will be created?
• Where will the data come from?
• Where will the data be held?
• How will the data be kept secure?
• What level of privacy is required?
• How the data or information will be analysed, curated, and presented?
• Who will have access or rights?
• How will integrity be ensured?
• How will data quality be assured?
• What will happen to the data after the project is completed?

These questions prompt the project management team to be proactive about how data will be
used to support decision making and project controls. The project context will greatly influence
the approach. For example, most of the data to support the project in the Data Knowledge
scenario will exist within or will be generated by the organization itself. Whereas, in the Findef
scenario, they will need to source market and user data externally to help inform product design
and development decisions.

Once the project management team understands the data management requirements for the
project, they can then consider their options and develop the management approach. When
developing the management approach, they will need to factor the data management capability
of the project team and the organizations involved; this will comprise data literacy of key
decision makers such as members of the project board, data skills of the project manager and
project support, data access, and the provision of automation, data analytics or generative AI
technology to make use of the data.

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 41
5.3.2 Digital technologies
Digital technologies can help with both project management and project delivery. Technology to
support project management will mostly be in support of the data management approach
described in 5.3.1. Technology to support project delivery will depend on the delivery method
and may be specific to each stage and each work package within the project. For example, the
technology to support the work packages involved in design or the technology to support the
work packages involved in construction.

There may be some crossover between technology to support project management and
technology to support project delivery. For example, the use of Augmented Reality (AR)
headsets can be used in design but can also provide data for managing requests for change or
for capturing details for issues and risks.

The project management team will need to consider how the technology across the project will
be sourced, set up, integrated, maintained, used, and decommissioned.

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 42
Chapter 6
Recommendations for
practice success

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 43
6. Recommendations for practice success
6.1 Recommendations
The successful use of automation, data analytics, AI and digital technology on projects requires
changes at the individual, project, and organizational levels. It may also change the profession
of project management.

The following five recommendations cover these different levels:

PRINCE2
Recommendation Comments
Element

Individual Competitive advantage will come for People

develop your data literacy and skills those who make better decisions
based on insights gained from AI.
Knowing the provenance of your
data, the type of algorithms used,
and the type of metrics used in the
outputs is now a core skill for project
professionals
“In a world of change, the learners
shall inherit the earth, while the
learned shall find themselves
perfectly suited for a world that no
longer exists. As a leader, you must
embrace change and remain a
student, or find yourself perfectly
suited for a world that no longer
exists” Eric Hoffer

Project Be curious. How is AI being applied All practices

make deliberate and conscious to other parts of your work and life
decisions on how you use AI on and could that same approach work
your project. on your project?
Where and how can your project
benefit from increased use of
automation, data-analytics, and AI
on your project?
Capture these decisions in the
project’s digital and data
management approach

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 44
 PRINCE2
Recommendation Comments
Element

Project/organization Automating poor practices or poor Learning from

focus on data first, then systems processes may just make issues experience
happen faster. Making decisions principle
based on bad data is likely to result
in bad decisions. Therefore, the
primary focus needs to be on data.
What data do you need, where will it
come from, how will you use it and
keep it secure, what will happen to it
when the project completes

Project/organization Organizational design and People

development should cover capability
understand the digital/data Organizing
gaps in the project
capability you have and what you practice
need

Profession Projects are complex, involving lots All aspects

of uncertainty and ambiguity. Some
can data analytics and AI solve
elements of project management
persistent areas of under-
may be better understood through
performance that methods and
deep learning algorithms than what
models have struggled to address?
we have gained from research to
date. Some elements of project
management may be better
performed by AI than humans.

We all need to be ready for be ready

for exponential change (see below)

6.2 Be ready for exponential change: where next?

We live in exponential times where the rate of change is accelerating. Following a launch in
1999, Netflix took 3.5 years to achieve 1 million users. Spotify took 5 months from launch in
2008. ChatGPT took 5 days. Threads took 1 hour.

Azeem Azhar remarked that “if the primary cause of the exponential gap is our failure to predict
the cadence of exponential change, the secondary cause is our consequent failure to adapt to it.
As the speed of change increases, our society remoulds itself at a much slower pace. Our
institutions have an in-built tendency towards incrementalism”. In the last 30 years project
delivery has changed very little.

Project data is typically a by-product of the project delivery process, an exhaust plume. As
organizations begin to recognize its strategic importance the volume of project data properly
captured will grow exponentially. They will recognize that there are sources of data that are
being overlooked. Data that, when integrated, helps enhance AI model performance. The data

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 45
integration challenge will become exponentially more difficult, requiring cross-organizational
collaboration on a scale we have never imagined. The climate emergency places additional
pressure on all of us to work and think differently. Commercial competitive advantage will
become secondary to collegiate working, with major clients rewarding those organizations that
move the dial in the interests of society. Effective project delivery holds the key to unlocking
many of the solutions that will help us to get to net zero and improve climate resilience.

We are in the middle of a perfect storm. How will you and your organization respond?

It may be helpful to reflect on four possible futures, and what this means for you and your
organization:

1. Project professionals supported by data professionals: project professionals

continue along the current path and leverage a service from a separate data function to
create data pipelines, apps, and insights. The rate of progress will be constrained by the
vision, bandwidth, and appetite for transformational change, which will vary by
organization. The more progressive and visionary organizations will outperform their
peers.

2. Data professionals supported by project professionals: data professionals review

every part of the project delivery process and explore how to automate, integrate, and
accelerate it. The ultimate objective could be to create a fully automated project or get as
close to it as possible. Removing interfaces, automatic creation of schedules, automated
creation and review of documentation, designs, logistics optimization, and more. Even
stakeholder engagement can be hyper-personalized to an audience of one using LLM.
There will always be a human in the loop, but how close to full automation will they get?

3. A blended team: where project professionals acquire new skills, supported by a data
centre of excellence. Project professionals develop the knowledge to leverage tools to
free up capacity. This creates the bandwidth to engage with the data team to reimagine
how projects are delivered. They begin to run experiments and pilots at pace, feeding
back and iterating. This can be accelerated x100 by collaborating across organizational
boundaries. Rather than pursuing a headcount/cost reduction methodology, they pursue
approaches that deliver a step change in project delivery performance. Better insights,
better decisions, better projects, better outcomes.

4. The bots take over: the rate of progress is dictated by the vendors. New capabilities
emerge from OpenAI, Microsoft, Oracle, and others that challenge the fundamentals.
They target the very essence of project delivery, leveraging vast amounts of data that
already exists within the cloud. Imagine a capability such as Midjourney, where the
images become better and better as humans engage with it and help it to learn. The
same could apply to project management.

Each of these futures is possible and probable.

6.2.1 Individual/personal perspective

The challenge for all of us is whether project data analytics will become a function of the IT and
data team, or an integral part of project delivery roles.

We see a range of responses. Some people who are curious but rely on others to find ways of
applying automation, data analytics and AI on their project (whether from within the project team

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 46
or from a central IT or data team). Other people who have acquired new skills who have been
promoted into new roles that simply did not exist a year ago. They are developing new products
and services for their organization, driving up delivery performance and unlocking new ways of
working. There are also people who are acquiring new skills, but their organization does not yet
have the infrastructure and systems in place to enable them to flourish.

Our advice is simple:

1. acknowledge that we are now in a new era
2. embrace the opportunities to free you and your colleagues from repetitive and
burdensome work
3. use this additional capacity to acquire new skills, immerse yourself in understanding
the art of the possible
4. evaluate whether you would like to become a practitioner or a translator; someone
who understands how the capabilities work, how they can be leveraged and forge a
bridge between data, IT, and project teams to enable change
5. pilot, be bold and experiment, use these experiments to understand your data,
constraints, obstacles and who the advocates are. If you do not have the support you
need, consider a move to somewhere that embraces data-driven project delivery.

What could this mean for project delivery roles? The future is unknown, but we offer some
possible developments.

• Risk managers: we tend to view risk management as a process, underpinned by a

range of different systems, risk registers and tools. Ultimately, we are guessing what
may go wrong, quantify this guess, integrate it within a schedule, apply Monte Carlo
analysis then derive a very precise three-point estimate that becomes encoded in future
business cases. Imagine a future where we pivot towards variance analytics, where we
recognize that variance can emerge from risk, change control, technical queries, and a
multitude of other sources. If we acquire a deeper understanding of the extent to which
events, work packages, or activities are predisposed to variance, we can target our
effort. As we increasingly understand causality, we can deploy probes and identify early
warnings to capture this variance earlier. We also use these methods to tailor our
approach to whether we are working within an ordered or complex adaptive system. It
could become a very different role.
• Stakeholder managers: we often think of stakeholder management in the context of
power and influence grids, which are often static documents. If we pivot our thinking
around influencers, we begin to target the small percentage of people who have the
large impact on outcomes. We can measure this in near-real time. We can use LLMs to
tailor messaging to specific stakeholder groups, then tailor again based upon any
previous communication or their perceived level of support. We see our role as
influencers rather than trying to manage stakeholders. Many of them simply cannot be
managed, but we can get better at understanding them and trying to win them over.
• Benefits managers: this will no longer be a process-centred around benefits mapping.
Those who are proficient in data will be adept at understanding the decision trade space,
modelling scenarios, and informing decisions at key points. Benefits will become more
extensive, linked to the UN Sustainable Development Goals and beyond. Benefits
managers will understand how decisions impact these goals or enhance return on
investment in the broadest sense. Those who can leverage insights from their peers will
ultimately deliver better performing models, augmenting human intuition. They will be
data driven, rather than process driven.

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 47
 • Project managers: there is an opportunity to pivot the approach away from process
centric outputs to real time insights, fed by a steady stream of validated and integrated
data. The project managers who can see around corners, identify emerging issues much
earlier, and leverage the hard-won experience from other projects will outperform those
who do not. They unlock new superpowers which would otherwise not exist. They will
also have the data to challenge intrinsic bias, encouraging key stakeholders to reflect on
their position or decisions.

This does not just require people just to learn how to use new tools or wrangle data, but to
unleash the art of the possible.

6.2.2 Organizational perspective

The organizational perspective is inextricably linked to ambition. To what extent does the
organization want to move the dial on how it delivers projects? Is project data analytics a bolt on
established ways of working or an integral part of reimagining project delivery.

The approach will also change based upon organization type. Will advanced data analytics
provide your organization with a competitive advantage? Will you use the data to develop new
tools that provide part of your SAAS offering? Will those who hoard the most data ultimately
have better trained models? Or are you a public client that seeks to extract more value from
each dollar of investment by leveraging the hard-won experience of all those projects that have
gone before, encouraging competition based on innovation rather than data hoarding or narrow
commercial interests? These perspectives will often be in conflict, requiring clients to establish
new ways of working and reward those who are collegiate, building on the shoulders of those
who have gone before.

Another perspective is organizational design. Is data an enabling service or an integrated part of

everything that your organization does? How will it be used to reimagine project delivery and
drive a step change in performance? This will drive the organizational design. From the level of
investment through to governance, data integration, data access, new methods and tools, real
time insights, and much more. It depends whether you are a tier 2 contractor within a broader
ecosystem or a dominant tier 1 contractor who can dictate the rules.

6.2.3 Project management profession perspective

Project data analytics will challenge the very essence of a body of knowledge. A compendium of
good practice that has evolved over the years. As data volumes grow, we will be able to
evidence which practices and methods work best in which situations. The insights will become
dynamic and adapt to context. We can train LLMs on not only the methods and processes, but
all the variance events, design queries, trade-off decisions, norms, and much more. We move
beyond an instruction manual to a fully integrated project assistant.

LLMs need large volumes of data to be trained on. A challenge for the project management
profession is how we access the data to train them as project data tends to be siloed or not
maintained due to the temporary nature of projects. We see several scenarios:
1. The big tech firms such as Google, Microsoft, and Amazon develop project
management solutions based on the data they harvest.
2. Tool vendors use your data through permissions you agree to in the licence
agreement. They train their tools with the data they have (including yours) and then
charge their clients for the insights gained through further subscriptions or use the
insights for competitive advantage to gain new clients and retain existing ones.

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 48
 3. Companies establish their own LLMs based on their own project data. Likely to be
viable only for larger organizations with large volumes of projects.
4. Data trusts are set up for industry sectors, possibly by industry or trade associations,
where they pool their data across multiple organizations with specific arrangements
for how data is cleaned, anonymised, secured, and accessed. There are several
data trusts in existence to date, for example the Construction Data Trust in the UK.
5. National governments establish data lakes from the data generated from publicly
funded projects through legislation, and then make the insights or data available to
different degrees depending on the organization requesting access to it. This is
already happening to some extent, but there are limitations on how far into the
supply chain the governments can get the data from.
6. Professional bodies and other IP owners develop their bodies of knowledge using
data from a combination of the above.

What underpins all these future possibilities is challenge and opportunity. The opportunity for
professionals to reskill and upskill and for organizations to gain competitive advantage by
leveraging technology alongside proven best practice to improve project success. As the world
around us evolves, we too must adapt, innovate, and stay at the forefront.

Your opinion matters!

We invite you to take a few minutes to give us your feedback about the PRINCE2 7 AI
Practice Guide by completing our online Feedback Form here.

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 49
 PeopleCert is accredited by Lloyd’s (UK) in accordance
with ISO 14001 for Environmental Management, since
ESG
2006. Recognized through numerous awards, we
Commitment
remain committed to ESG leadership and the
preservation of our planet.

PRINCE2® 7 AI Practice Guide | Global Best Practice | Copyright© PeopleCert International Ltd. | 50