Artificial Intelligence, Privacy and Ethics

What is Artificial Intelligence?

Artificial Intelligence The replication of the human ability to perform cognitive tasks such as
Perception, Anticipation, Decision Making and other aspects of Executive Function in Machines
and Computer Systems. Machine Learning is an approach to achieving Artificial Intelligence in
which large amounts of data is parsed to the selected algorithms so that they are trained to
make a determination or prediction about something in the world. Today Machine Learning
exists in mainly three forms: Supervised Learning, Unsupervised Learning and Reinforcement
Learning.

In Supervised Learning, the selected Machine Learning Model is fed with data and labels that
correspond to the data (email is a common example). The goal is to learn a general rule that
maps inputs to outputs.

In Unsupervised Learning our Machine Learning Model is given unlabeled data and the task of
structuring the data. Unsupervised Learning can be used for discovering hidden patterns,
finding groupings and inference. Customer segmentation is a common example.

In Reinforcement Learning, an Agent, interacts with an Environment with the aim of achieving
a particular goal. The agent learns based on the outcome or reward of the actions taken.
These rewards can be positive or negative. Autonomous driving is an example.

The use of Machine Learning is commonly used for situations where deterministic programming
will not adequately be able to model the problem.

Machine Learning is often found in environments where:

1. Problems where there are no human experts, so data cannot be labelled or
categorised;
2. Problems where there are human experts, but it is very hard to program;
3. Problems where there are human experts and it can be programmed, but where it is
not cost effective to implement.
Common Applications of Artificial Intelligence
Due to the explosion in computing power that we have seen in recent years, Artificial
Intelligence has become far more practical and widespread in its application to problems faced
in many industries. It should also be noted that the increasing amounts of data and company's
desire for competitive advantage also fuels the advance of artificial intelligence. As Globalization
continues to increase, so will the proliferation of Artificial Intelligence.

Today Artificial Intelligence is used in the following industries:

Healthcare: In recently years we have seen companies such as IBM (Using Watson)
and Google DeepMind make advances in their applications Artificial Intelligence in the
field of medicine. IBMs Watson can diagnose a number of diseases and prescribe
treatment based on what it would have learned from the cases it would have analyzed.
DeepMind is also aiming to revolutionize the Healthcare process itself by enabling more
staff to know the history of patients and understand the individual need of its patients.

Customer Service: Today we see the increased use of chatbots such built on various
platforms such as Facebook and Slack. The customer service industry has been
revolutionized and is on a trajectory where we are able to see personalized customer
service being offered at scale (and reduced cost) in a variety of industries at virtually any
time of day.

Education: Education is a billion-dollar industry contributing heavily to the various

economies. EdTech, as we all know, is a booming industry. AI could be used to grade
papers, bots to answer questions, personal assistant systems that act as tutors, virtual
reality for hands-on learning, etc.

Finance: The financial data has been increasing at a rapid rate making it harder for the
financial services companies to keep up. With the help of Artificial Intelligence, predictive
systems are being put in place to forecast stock trends and manage finances. Recently,
we have also seen the trend of Robo-Advisors that are automating the industry.

In each of the industries mentioned above, Artificial Intelligence and Machine Learning is being
used to improve the quality of service and the products offered to consumers. The aim is to
understand and learn more about the consumer. The aim is to understand their goals, their
lives, their challenges and their contexts. Artificial Intelligence allows companies to harness the
data they gather on customers so that they can innovate and make their products even more
relevant and appropriate to the needs of their customers.

In todays landscape we have seen companies such as Amazon and Google make use of
Artificial Intelligence to gain even further access into our lives. Products such as Google Home,
and Amazon Echo have created avenues for those companies to learn even more about us as
they leverage advanced artificial intelligence techniques to learn about us on a deeper and more
intimate level.

Artificial Intelligence not only decreases costs but also saves time, increases accuracy and
productivity. Google estimates that robots will reach levels of human intelligence by 2015, one-
third of jobs will be replaced by robots and other smart machines.

Companies such as these are able to use Artificial Intelligence in the following ways:

Speech Recognition
In the 1990s, computer speech recognition reached a practical level for limited purposes. Thus
United Airlines has replaced its keyboard tree for flight information by a system using speech
recognition of flight numbers and city names. [Google home can now distinguish users]

Natural Language Processing

The computer has to be provided with an understanding of the domain the text is about, and this
is presently possible only for very limited domains.

Computer Vision
The world is composed of three-dimensional objects, but the inputs to the human eye and
computers' TV cameras are two dimensional. Some useful programs can work solely in two
dimensions, but full computer vision requires partial three-dimensional information that is not
just a set of two-dimensional views.

Expert Systems
A ``knowledge engineer'' interviews experts in a certain domain and tries to embody their
knowledge in a computer program for carrying out some task. How well this works depends on
whether the intellectual mechanisms required for the task are within the present state of AI.
heuristic classification
One of the most feasible kinds of expert system given the present knowledge of AI is to put
some information in one of a fixed set of categories using several sources of information. An
example is advising whether to accept a proposed credit card purchase.

AI in the InfoSec landscape

As more and more data is gathered on users to enable product innovation, companies will
become an even greater target for cybercrime. Tripwires Irfan Khimji, wrote that Cybercrime Is
Now More Profitable Than The Drug Trade stating that those who commit cyber crime have
begun targeting customers payment information, in particular credit card information. Any knd
of data that has commercial value should be considered as a target. As Artificial Intelligence
continues to advance, we can expect to see far more elaborate attacks than we see today. One
of the greatest threats is the emergence of intelligent, self-learning, self-augmenting computer
viruses.

Artificial Intelligence can be used to execute and defend against cyber attacks. Companies such
as Darktrace have created an AI that can be used to analyze a networks traffic and learn what
normal traffic looks like over a period of time. After learning what the network profile, it can
detect any anomalous activity coming from any node on the network and take action.

Conventional approaches to cyber security have relied on signatures and rule-based logic in
order to guard against known, understood and documented threats. Which means that action
can only be taken after a solution to a new threat has been found. This of course relies on the
fact that the threat has been detected in the first place. With attacks becoming increasingly
sophisticated, this can take months to happen. Studies have shown that in 2016, it took an
average of 99 days before infiltration by cyber threats could be detected. Even though it is a
marked improvement from the 146 days it took in 2015, that time frame gives attackers a
significant amount of time to find, steal, learn about and manipulate a companys data.

Even a few years ago, launching a distributed denial of service attack to take down a website,
defacing webpages and stealing credit card details were considered major instances of cyber-
attacks. Additionally, these systems could work simultaneously on various tasks, monitoring and
protecting a vast number of devices and systems. They can, therefore, mitigate large scale
attacks.

More worryingly, he said, cyber-attackers are focused on not only stealing the data but also
altering them without detection. If an attacker can alter a single row or column of data in a
database once a month, undetected, in the long run the consequences can be severe because
companies would find it hard to distinguish between real and fake. For example, if electronic
medical records are altered without the knowledge of doctors and nurses, it could potentially
lead to misdiagnoses that could put patients' lives in danger.

Conclusion
Artificial Intelligence is being used in the world of business to enrich the products and services
that are being delivered to consumers. Added to which, with the advance of globalization, the
number of people and systems that have been connected together has grown exponentially.
This means that in the InfoSec landscape today, Cyber attacks, can have a global impact.
Traditional approaches are quickly fading into obsolescence due to the fact that they are not
able to provide a response quickly enough to todays new and intelligent threats. Artificial
Intelligence has now become necessary for combating these new threats. In todays InfoSec
landscape, Artificial Intelligence can be both a friend and a foe.

References
https://www.theglobeandmail.com/technology/dear-valued-customer-thank-you-for-giving-us-all-
your-personal-data/article582786/
https://medium.com/startup-grind/artificial-intelligence-is-taking-over-privacy-is-gone-
d9eb131d6eca
http://resources.infosecinstitute.com/artificial-intelligence-impact-cyber-security/
https://hbr.org/2017/05/ai-is-the-future-of-cybersecurity-for-better-and-for-worse *
https://www.forbes.com/sites/forbestechcouncil/2017/05/23/is-cybersecurity-a-second-coming-
for-ai/#cb3db957c400
http://www.kdnuggets.com/2017/06/kaspersky-artificial-intelligence-bubble-future-
cybersecurity.html
http://www.cnbc.com/2017/04/17/darktrace-on-why-artificial-intelligence-is-key-in-
cybersecurity.html
http://www.darkreading.com/threat-intelligence/artificial-intelligence-cybersecurity-friend-or-foe-
/a/d-id/1328838
https://www.forbes.com/sites/forbestechcouncil/2017/05/23/is-cybersecurity-a-second-coming-
for-ai/#6e3346227c40
https://www.weforum.org/agenda/2016/10/top-10-ethical-issues-in-artificial-intelligence/
http://fortune.com/2017/02/06/artificial-intelligence-ethics-disasters/
http://www.zdnet.com/article/artificial-intelligence-legal-ethical-and-policy-issues/
http://www.bbc.com/future/story/20170307-the-ethical-challenge-facing-artificial-intelligence
https://www.bloomberg.com/news/articles/2017-06-23/google-will-stop-reading-your-emails-for-
gmail-ads
https://techcrunch.com/2017/04/20/google-home-can-now-recognize-up-to-six-voices-and-give-
personalized-responses/
https://www.tripwire.com/state-of-security/regulatory-compliance/pci/cybercrime-is-now-more-
profitable-than-the-drug-trade/