Compliance and Whistleblowing
Compliance and Whistleblowing
A. INTRODUCTION
Whistleblowing is not a new phenomenon. Some scholars have traced the 15.001
concept to Ancient Greece, drawing parallels with the notion of parrhēsia,
or fearless speech.2 Lykourgos, an Athenian orator of the mid-300s BC, is
1 The authors acknowledge with thanks the helpful comments of Ashley Savage and Jelena Madir on an earlier
draft. Thanks are also owed to Maria Shepard and Emma Franklin for their assistance with research for the
second edition of this chapter.
2 Michel Focault, Fearless Speech, Semiotext(e) (2001); and Alan Chu, In Tradition of Speaking Fearlessly:
Locating a Rhetoric of Whistleblowing in the Parrhēsiastic Dialectic, 19 Advances in the History of Rhetoric
231 (2016), at 239–48.
365
reported to have said that ‘neither the laws nor judges can bring any results
unless someone denounces the wrong doers’.3 Laws to incentivise whistle-
blowers are not novel either. In the 7th century, a British king declared that
‘if a freeman works during [the Sabbath], he shall forfeit his [profits], and
the man who informs against him shall have half the fine, and [the profits]
of the labour’.4 Modern American whistleblower protections, meanwhile, are
grounded in the Civil War-era False Claims Act.
15.002 The label whistleblowing, on the other hand, is a more recent invention. The
term was popularised in the 1970s by American political activist Ralph Nader,
who described it as ‘an act of a man or a woman who, believing that the public
interest overrides the interest of the organisation he serves, publicly “blows the
whistle” if the organisation is involved in corrupt, illegal, fraudulent or harmful
activity’.5 In the following decades, whistleblowing entered the mainstream
lexicon. High-profile whistleblowers drew attention to the considerable public
interest in their deeds and the adverse consequences they often suffered. Laws
were enacted to encourage whistleblowing and protect those who did so – the
Public Interest Disclosure Act 1998, a notable early example in the UK – and
charitable organisations were established to advocate the whistleblower cause.
There remains, though, no universally-accepted definition of a whistleblower
or defined criteria of what constitutes whistleblowing.
366
among the first to raise the alarm about the gravity of COVID-19.7 Since then,
whistleblowers have exposed unsafe work conditions in hospitals, schools, and
businesses, abuses of privacy, government misspending and other wrongdoing.
The intimidation, lawsuits and job-loss that many faced as a result has led to
renewed calls for global authorities to strengthen whistleblower protections.8
In recent years Ireland, the Netherlands, France, Italy and Serbia have been
among the jurisdictions to pass landmark whistleblower protection regimes –
just under 50 countries globally now have specific legal protections for those
who blow the whistle. In 2019, the European Union passed a landmark whis-
tleblower protection directive requiring all member states to introduce best
practice laws in the coming years.
But just as societies are beginning to appreciate the significant contributions 15.004
whistleblowers make to democratic accountability and corporate compliance,
the concept of whistleblowing is being transformed by technology. Such
disruption is the focus of this chapter. Technology offers much promise to
whistleblowers and whistleblower protections, but also many potential pitfalls.
A sober analysis is required to determine where technology might add benefit,
and where it could prove problematic.
This chapter has three substantive parts. Section B will begin by detailing 15.005
how data analytics and artificial intelligence (AI) are becoming ‘algorithmic
whistleblowers’, detecting (or even preventing) misconduct before it can be
discovered by human whistleblowers. In other words, how is technology
replacing whistleblowers? Section C analyses a range of technological solu-
tions that are helping to empower whistleblowers, by protecting them and
giving them new avenues for reporting misconduct. The exciting potential for
blockchain to offer anonymity, immutability, resilience, compensation and
information escrow in the whistleblowing context will be considered. Section
D discusses how technology is changing the nature of whistleblowing, given
the modern-day ability to distribute terabytes of information with a few clicks.
The implications of technological-driven change, including the increasingly
7 Nie, Jing-Bao, and Carl Elliott. Humiliating Whistle-Blowers: Li Wenliang, the Response to Covid-19,
and the Call for a Decent Society. Journal of Bioethical Inquiry, 1–5. 25 Aug. 2020, doi:10.1007/s11673-020
-09990-x.
8 International Bar Association, Authorities Urged to Protect Whistleblowers during Covid-19 crisis
(2020), available at: https://www.ibanet.org/Article/NewDetail.aspx?ArticleUid=76b74307-2379-4f3d
-93b1-a8d2b963338b; Transparency International, governments and corporations need to guarantee safety
of covid-19 whistleblowers (2020), available at: https://www.transparency.org/en/press/governments-and
-corporations-need-to-guarantee-safety-of-covid-19-whistleblowers.
367
blurred lines between whistleblowing, leaking and hacking, are reflected upon.
Section E concludes.
15.007 Transparency is a contested concept, but at its heart it ‘refers to the notion
that information about an individual or organisation’s actions can be seen
from the outside’.11 Transparency has become a default policy prescription,12
often invoked as an essential component of trust and cooperation, as a market
efficiency mechanism, as a legitimising procedural tool and, at its broadest, as
a value embedded in democracy.
15.008 The ascent of transparency as an institutional norm dovetails with the growing
recognition of the value of whistleblowers as ‘the primary source of involuntary
transparency’,13 reflected in the development of whistleblower protection leg-
islation worldwide.14
9 L. Carolan, Open data, transparency and accountability: Topic guide (2016), available at: https://assets
.publishing.service.gov.uk/media/5857fdcb40f0b60e4a0000d6/OpenDataTA_GSDRC.pdf at 5.
10 Adam Waytz, Why Robots Could be Awesome Whistleblowers (2014), available at: https://www.theatlantic
.com/business/archive/2014/10/why-robots-could-be-awesome-whistleblowers/381216/.
11 Matthew S. Mayernik, Open Data: Accountability and Transparency, Big Data and Society 1 (2017), available
at: https://doi.org/10.1177/2053951717718853, at 1.
12 Aarti Gupta, Transparency Under Scrutiny: Information Disclosure in Global Environmental Governance, 8
Global Environmental Politics 1 (2008), at 1.
13 Jennifer Shkabatur, Transparency With(out) Accountability: Open Government in the United States 31
Policy Review 89 (2012), available at: https://digitalcommons.law.yale.edu/cgi/, at 113.
14 International Bar Association, supra note 4.
368
Meanwhile, regulators and advocates alike are increasingly pursuing another 15.009
transparency frontier which has, along with whistleblowing, grown from the
historical right to information movement: access to data.15 Open data propo-
nents support the disclosure of data in a way that allows it to be freely used,
modified and shared by anyone for any purpose. The OECD identifies open
data as a ‘key public good’ and a powerful tool in the fight against the abuse
of power.16 Efforts are being made to open up government data sets which
include public officials’ directories, budgets, public procurement, political
financing, voting records and land registries.17
The ‘openness revolution’ is also marching into the private sector.18 For 15.010
example, the movement pushing for a global public database featuring
country-by-country reporting (CBCR)19 on the economic activity and tax
contributions of multinational corporations achieved a breakthrough in 2017,
when the European Commission voted for the second time in favour of
public CBCR by multinationals.20 Another example is the call for beneficial
ownership reporting.21 In 2016, the UK became the first country to publish
the identity of those who benefit from, own and control companies;22 and
in April 2020, 18 countries were found to have public beneficial ownership
registers.23 Meanwhile, navigating the regulatory complexity that followed the
Global Financial Crisis has ‘inevitably required greater granularity, precision
and frequency in data reporting, aggregation, and analysis’ from corporations,
15 Katleen Janssen, Open Government Data: Right to Information 2.0 or its Rollback Version? 8 ICRI
Research Paper (2012), available at: https://ssrn.com/abstract=2152566, at 4–8.
16 OECD, Compendium of Good Practices on the use of Open Data for Anti-corruption (2017), available at:
http://www.oecd.org/gov/digital-government/g20-oecd-compendium.pdf.
17 World Wide Web Foundation and Transparency International, Connecting the Dots: Building a Case for Open
Data to Fight Corruption (2017), available at: http://webfoundation.org/docs/2017/04/2017_OpenDataConn
ectingDots_EN-6.pdf.
18 The Openness Revolution (2014), The Economist, available at: www.economist.com/business/2014/12/11/
the-openness-revolution.
19 Since 2002, CBCR has become the extractive industry standard in more than countries, and has since
spread to the financial institutions. See Alex Cobham et al., What Do They Pay? Towards a Public Database
to Account for the Economic Activities and Tax Contributions of Multinational Corporations (2017), available at:
datafortaxjustice.net/what-do-they-pay/#extractive-industries-data.
20 Financial Transparency Coalition, Letting the Public In (2015), available at: https://financialtransparency
.org/wp-content/uploads/2016/09/OpenData_fullpaper.pdf.
21 According to the World Bank, up to 70 per cent of cases of financial misconduct involve anonymous compa-
nies. Open Ownership: Ending anonymous company ownership, available at: https://openownership.org/.
22 Jonathan Grey and Timothy Glyn Davies, Fighting Phantom Firms in the UK: From Opening Up Datasets
to Reshaping Data Infrastructures? (2015), available at: doi:10.2139/ssrn.2610937.
23 U4 Anti-Corruption Resource Centre and Transparency International, Beneficial ownership registers:
Progress to date, available at: https://knowledgehub.transparency.org/assets/uploads/helpdesk/Beneficial
-ownership-registers_2020_PR.pdf annex 1, at 17.
369
15.011 In summary, the demand for open data coupled with an increase in data
intensive regulation is adding unprecedented dimensionality to high-volume,
high-velocity and high-variety information assets, also known as big data.26 At
the outset, this might appear to be an unreservedly good development from
transparency and anti-corruption perspectives, reducing reliance on human
whistleblowers. However, there are at least three potential limitations.
2. Challenges
15.012 First, more data does not necessarily mean more transparency. Jonathan
Fox distinguishes between two kinds of transparency.27 Opaque transpar-
ency involves ‘the dissemination of information that does not reveal how
institutions actually behave in practice, whether in terms of how they make
decisions, or the results of their actions’, while clear transparency ‘sheds light
on institutional behaviour permit[ting] interested parties to pursue strategies
of constructive change’. Despite a tendency to equate more data with more
transparency, clear transparency necessitates not data per se, but the ability to
extract relevant information about the entity in question from that data.28 Even
putting aside data quality issues,29 the sheer volume of potentially available
data30 and a dearth of data literacy31 among the general population makes
actualisation of the average citizen as auditor doubtful.
24 Douglas W. Arner et al., FinTech, RegTech, and the Reconceptualization of Financial Regulation 37(3)
Northwestern Journal of International Law and Business (2017), available at: https://scholarlycommons.law
.northwestern.edu/njilb/vol37/iss3/2, at 388.
25 Bob can der Made, European Union: The Revival of Public CbCR Amid New interest in ESG Transparency
(2020), available at: https://www.internationaltaxreview.com/article/b1kzwk564ld7k3/european-union-the
-revival-of-public-cbcr-amid-new-interest-in-esg-transparency.
26 Doug Laney, 3D Management: Controlling Data Volume, Velocity, and Variety, Gartner (2001), available
at: https://blogs.gartner.com/doug-laney/files/2012/01/.
27 Jonathan Fox, The Uncertain Relationship between Transparency and Accountability, 663 Development in
Practice (2007), available at: https://doi.org/10.1080/09614520701469955, at 667.
28 Catharina Lindstedt and Daniel Naurin, Transparency is not Enough: Making Transparency Effective in
Reducing Corruption, International Political Science Review (2010), available at: https://journals.sagepub
.com/doi/abs/10.1177/0192512110377602, at 302.
29 Open Knowledge International Blog: Open Data Quality – the Next Shift in Open Data? (2017), available at:
https://blog.okfn.org/2017/05/31/open-data-quality-the-next-shift-in-open-data/.
30 Of all data existing in 2018, 90 per cent was created in 2016–2018, amounting to 2.5 quintillion bytes of data
created per day. See Domo, Data Never Sleeps 5.0 (2018), available at: https://www.domo.com/learn/data
-never-sleeps-5.
31 Annika Woolf et al., Creating an Understanding of Data Literacy for a Data-driven Society, 12 Journal of
Community Informatics (2016), available at: http://oro.open.ac.uk/47779/, at 10.
370
Second, the push for transparency has resulted in a fragmented web of finan- 15.013
cial regulations, contributing to ever-increasing compliance costs. The rate of
new regulation led one analyst to suggest that ‘much like Moore’s law in the
field of computing there is a “Regulatory Law” that means the operational
burden of controlling regulations will double every few years’.32 Third, reg-
ulators are under considerable pressure to effectively supervise with limited
resources, even as technology is enabling innovative difficult-to-trace methods
for abusing power. For now, reliance on whistleblowers persists. Disruptive
technology might provide opportunities for these concerns to be addressed.
3. Opportunity
AI has been taken to include machines that exhibit aspects of human intelli- 15.015
gence like problem solving, making predictions, identifying objects and ana-
lysing language.34 Machine learning is one subset of AI. Supervised machine
learning algorithms are ‘trained’ through the processing of labelled samples of
training data by a learning algorithm, before the algorithm is presented with
unlabelled test data. Typical applications include the prediction of a label (clas-
sification) or a continuous value (regression). Unsupervised learning involves
tasks like clustering and dimensionality reduction, in order to ‘learn the inher-
ent structure of our data without using explicitly-provided labels’.35
Deep learning is an approach to machine learning which departs from the 15.016
statistics-based methods that ground the solutions previously described. Deep
32 Tom Groenfeldt, Taming The High Costs Of Compliance With Tech (2018), available at: www.forbes.com/
sites/tomgroenfeldt/2018/03/22/taming-the-high-costs-of-compliance-with-tech/#3f7d5285d3f7.
33 Tom Simonite, How can AI keep Accelerating after Moore’s Law (2017), available at: https:// www
.technologyreview.com/s/607917/how-ai-can-keep-accelerating-after-moores-law/.
34 While beyond the scope of this chapter, the definition and meaning of ‘artificial intelligence’ is fiercely con-
tested. Shane Legg and Marcus Hutter, A Collection of Definitions of Artificial Intelligence, 157 Frontiers
in Artificial Intelligence Appl. 17 (2007), available at: https://arxiv.org/pdf/0706.3639.pdf.
35 Devin Soni, Supervised vs. Unsupervised Learning, Towards Data Science (2018), available at: science.com/
supervised-vs-unsupervised-learning-14f68e32ea8d.
371
learning algorithms learn via layers of artificial neural networks imitating the
biological structure and functions of the brain.36 Whereas the performance of
trained machine learning algorithms will at some point reach a plateau, the
ability of deep neural networks to replicate real world systems has no such
theoretical ceiling.37 Most promising is deep learning’s superior potential to
discover structures within otherwise unstructured, unlabelled data – the format
of most data in the world.
15.018 Within companies, RegTech can prevent and detect asset misappropriations,
corrupt schemes and financial statement fraud, which would be otherwise
indiscernible to a human analyst.39 AI is also enabling auditors to analyse data
and detect connections between e-mails, pdf documents, expense reporting,
social media profiles, criminal record checks, work hour reports, registered
attempts to access restricted work areas and more.40 This could even reveal
behavioural insights so that ‘companies can identify individuals who might
pose a higher risk to business’.41
36 Snezana Agatonovic-Kustrin and Roderic Beresford, Basic Concepts of Artificial Neural Network (ANN)
Modeling and its Application in Pharmaceutical Research, 22 J Pharm Biomed Anal 171 (2000) available at:
https://doi.org/10.1016/S0731-7085(99)00272-1, at 718–22.
37 Ian Goodfellow et al., Deep Learning, The MIT Press (2016), at 197.
38 See generally: Financial Stability Board, The Use of Supervisory and Regulatory Technology by Authorities
and Regulated Institutions (2020) https://www.fsb.org/wp-content/uploads/P091020.pdf.
39 Institute of International Finance, Deploying Regtech Against Financial Crime (2017), available at: https://
www.iif.com/system/files/32370132_aml_final_id.pdf.
40 Ibid.
41 Ibid.
372
42 The National Bank of Rwanda (BNR) was one of the first financial institutions to implement this. Bank for
International Settlements, Innovative Technology in Financial Supervision (Suptech) – the experience of
early users (2018), available at: https://www.bis.org/fsi/publ/insights9.htm, at 6.
43 Business Reporting Language, or XBRL, is the international data standard for international business report-
ing: Marc D. Joffe, Open Data for Financial Reporting, Data Foundation (2017), available at: https://www
.datafoundation.org/xbrl-report-2017/.
44 Bank for International Settlements, Identifying regions at risk with Google Trends: the impact of Covid-19 on US
labour markets (2020), available at: https://www.bis.org/publ/bisbull08.pdf.
373
15.022 Although not typically conceptualised in this context, arguably SupTech can
also be used to refer to the use of technology by governments to detect and
deter systemic risks within their own agencies and to assist with the super-
visory activities of the public and independent watchdog organisations.47
One example is ProZorro, the much-lauded Ukrainian public e-procurement
system, which is being enhanced by AI to identify procurement violations and
tenders with a high risk of corruption, including COVID-19 related purchas-
es.48 Unlike orthodox risk management systems, the indicators are not pre-set
beforehand and there is no exhaustive list.49 In Colombia, the comptroller
general’s Oceano programme – another fraud detection analytics platform that
mines public procurement documents – recently blew the whistle on suspicious
links between companies and politicians related to emergency health spend-
ing.50 On 9 December 2020, in conjunction with the UN’s Anti-Corruption
Day, Microsoft unveiled its Anti-Corruption Technology and Solutions
(ACTS) which will reportedly work with governments and other organisations
to leverage Microsoft’s AI, cloud computing, and data visualisation technol-
ogies to ‘aggregate and analyse...enormous datasets in the cloud, ferreting out
corruption from the shadows where it lives, and even preventing corruption
before it happens’.51 The stated aim is to ‘help governments innovate’ and
45 Financial Stability Board, The Use of Supervisory and Regulatory Technology by Authorities and Regulated
Institutions Market developments and financial stability implications (2020), available at: https://www.fsb.org/
wp-content/uploads/P091020.pdf, at 59.
46 Bank for International Settlements, supra note 42.
47 Global Witness, Three Ways the UK’s Register of the Real Owners of Companies Is Already Proving
Its Worth (2018), available at: https://www.globalwitness.org/en/blog/three-ways-uks-register-real-owners
-companies-already-proving-its-worth/.
48 Transparency International, Where do we go from here to stop the pandemic? (2020), available at: https://
www.transparency.org/en/news/where-do-we-go-from-here-to-stop-the-pandemic.
49 Transparency International Ukraine, Dozorro Artificial Intelligence to Find Violations in ProZorro: How it
Works (2018), available at: https://ti-ukraine.org/en/news/dozorro-artificial-intelligence-to-find-violations
-in-prozorro-how-it-works/.
50 World Economic Forum, Why data is Latin America's best weapon against COVID-19 corruption (2020),
available at: https://www.weforum.org/agenda/2020/08/why-data-is-latin-americas-best-weapon-in-the
-fight-against-covid-19-corruption/.
51 Microsoft, Microsoft launches Anti-Corruption Technology and Solutions (2020), available at: https://blogs
.microsoft.com/on-the-issues/2020/12/09/microsoft-anti-corruption-technology-solutions-acts /.
374
eventually ‘bring the most promising solutions to the broadest possible audi-
ence’, which so far has included partnering with the IDB Transparency fund to
bring transparency to the use of COVID-19 stimulus funds.52
4. Obstacles
Various obstacles to RegTech and SupTech adoption remind us that techno- 15.023
logical solutions are no panacea for eliminating the challenges faced by whis-
tleblowers. As described in more detail in Chapter 12, these include regulatory
and legislative barriers to knowledge sharing (such as data protection and
localisation laws), legacy IT systems, the lack of integrated data taxonomies
and the limited room for financial institutions to innovate while maintaining
compliance.53 There are also at least three potential ethical challenges arising
from the proliferation of technology in the present context that merit further
examination.
a. Privacy
The first obstacle is the balance between transparency and privacy. As Fox 15.024
quips: ‘One person’s transparency is another’s surveillance.’54 A survey con-
ducted by Ernst & Young revealed a ‘tension between opinions about what
channels companies should monitor and the types of surveillance that their
employees consider a violation of privacy’.55 The GDPR, as the global legal
standard for data protection and privacy, imposes various duties on data
controllers and data processors, including obligations to declare a lawful basis
for data collection and processing, and limitations on the export of personal
data outside the EU. Special attention to these provisions should be paid by
organisations that are effectively outsourcing their RegTech and SupTech
compliance solutions to third parties.
52 Note, however, that Microsoft itself was allegedly implicated in a bribery scheme in Hungary, paying the
US SEC 25 million to settle the investigation in 2018. Kyle Wiggers, Microsoft launches effort to fight
corruption with AI and other emerging technologies (2020), available at: https://venturebeat.com/2020/12/
09/microsoft-launches-effort-to-fight-corruption-with-ai-and-other-emerging-technologies/ .
53 Institute of International Finance, supra note 39.
54 Fox, supra note 27; and Privacy International, Fintech: Privacy and Identity in the New Data-Intensive
Financial Sector (2017), available at: https://privacyinternational.org/sites/default/files/2017-12/Fintech
%20report.pdf.
55 For example, around 65 per cent of respondents felt that e-mail and phone-call monitoring was a violation of
privacy. See EY Reporting, What should be Monitored? (2017) available at: https://www.ey.com/Publication/
vwLUAssets/, at 9.
375
b. Bias
15.025 The second obstacle concerns the risks of error, bias and the threat of
algorithmic discrimination.56 Machine learning algorithms will learn from
and perpetuate distortions in training data. Moreover, inherently algo-
rithms are optimised to achieve particular goals, which can lead to biased
decision-making. RegTech and SupTech are not immune. For example, fraud
detection algorithms have been shown to be biased against certain ethnic
minorities, immigrants and even against men.57 While extensive technical
research is being done on identifying and correcting bias in algorithms, others
are advocating algorithmic impact assessments and even making a business out
of algorithmic auditing.58 ORCAA, one such consultancy, assesses the quality
of training data, testing the algorithms’ design, implementation, execution
and ethical consequences, and offers training in algorithmic auditing. The
resultant seal is ‘like an organic sticker for algorithms’, on the basis that ‘the
food we eat has quality certifications. Why shouldn’t the algorithms that shape
our world?’59 In the meantime, the question arises: how much less biased than
a human does an algorithm have to be before we are willing to let it loose on
the work of whistleblowers?
c. Black box
15.026 Finally, the third obstacle is that machine learning algorithms tend to be
‘opaque in the sense that … rarely does one have any concrete sense of how
or why a particular classification has been arrived at from inputs’.60 This is
known as the explainability or black box problem, and it is particularly acute
in deep learning. Some argue that even the technologically-increased accuracy
of decisions does not compensate for the inability to explain the weighting of
decision-making factors and essentially fails to respect a subject’s dignity,61
offending one’s ‘right to an explanation’,62 and raising Kafkaesque concerns
56 Solon Barocas and Andrew Selbst, Big Data’s Disparate Impact, 104 California Law Review 671 (2016),
available at: http://www.californialawreview.org/wp-content/uploads/2016/06/2Barocas-Selbst.pdf.
57 Adeesh Goel, Algorithmic Bias: Challenges and Solutions (2017), available at: https://mse238blog.stanford
.edu/2017/08/adeesh/algorithmic-bias-challenges-and-solutions/.
58 FAT/ML: Principles for Accountable Algorithms and a Social Impact Statement for Algorithms, available
at: http://www.fatml.org/resources/principles-for-accountable-algorithms.
59 Katharine Schwabe: This logo is like an organic sticker for algorithms (2018), available at: https://www
.fastcompany.com/90172734/this-logo-is-like-an-organic-sticker-for-algorithms-that-arent-evil.
60 Jenna Burrell, How the Machine ‘Thinks’: Understanding Opacity in Machine Learning Algorithms, Big
Data and Society (2016), available at: https://journals.sagepub.com/doi/pdf/10.1177/2053951715622512, at
1.
61 Jeremy Waldron, How Law Protects Dignity, 71 Cambridge Law Journal 200 (2012), at 210.
62 Reuben Binns, Max Van Kleek, et al., ‘It’s Reducing a Human Being to a Percentage’; Perceptions of Justice
in Algorithmic Decisions (2018), available at: https://doi.org/10.1145/3173574.3173951.
376
for fair trial standards.63 There are a growing number of researchers, business
leaders and policy makers who are developing both technical solutions to
explainable AI (XAI) and corporate civil regulation for the development of
ethical AI.64 On the other hand, some counter that to the degree that AI
becomes ‘explainable’, bad actors may be able to adjust their behaviour to
‘game’ the system.65 As it stands, the paradox is that the increase in trans-
parency in the sense of information disclosure is dependent on an opaque
mechanism. Minimising harm to whistleblowers requires a trade-off with the
potential of harm to the subjects of inexplicable algorithmic outputs.
The barriers to blowing the whistle are widely known and have been exten- 15.028
sively analysed, the foremost being the fear of retaliation.66 The difficulty of
ensuring confidentiality and, in some cases, the anonymity of the whistle-
blower therefore looms large. A related problem is the utilisation of trusted
channels of reporting, which must be secure and effective. Technological
applications, designed to facilitate the whistleblowing process, offer potential
63 Council of Europe, Algorithms and Human Rights (2017), available at: http://rm.coe.int/algorithms-and
-human-rights-en-rev/16807956b5.
64 See the 2017 Asilomar principles, which has 1,273 AI/Robotics researchers as signatories. The Future of Life
Institute, The Asilomar AI Principles (2017), available at: https://futureoflife.org/ai-principles/?submitted=
1&cn-reloaded=1#confirmation.
65 Paul B. de Laat, Algorithmic Decision-making Based On Machine Learning from Big Data: Can
Transparency Restore Accountability? P.B. Philos. Technol. 17 (2017), available at: https://link.springer.com/
content/pdf/10.1007%2Fs13347-017-0293-z.pdf.
66 International Bar Association, supra note 4.
377
15.029 This section aims to examine four technologies which are running in parallel
to these trends to empower the whistleblower: hotline services and web portals,
mobile apps and blockchain.73
67 For a more specific discussion of the potential of technology for addressing harassment, see Emma Franklin
and Kieran Pender, Innovation-led cultural change: can technology effectively address workplace harassment?
(2020), available at: https://www.ibanet.org/Document/Default.aspx?DocumentUid=4c00afd9-53e7-4ad6
-8db0-c663c2f23f45.
68 For example, Got Ethics A/S, available at: https://www.gotethics.com/en/; Whispli, available at: https://
www.whispli.com/; NAVEX Global's Whistleblower Hotline, available at: https://www.navexglobal.com/
en-us/products/hotline-reporting-and-intake; Your Call, available at: https://www.whistleblowing.com
.au/solutions/?gclid=Cj0KCQiA8dH-BRD_ARIsAC24umYuPBhBvQqxswSWcg4F1hO9nsXDf3Tnz1
kiIEK6sJtxw3tTMVzKsRYaAnFYEALw_wcB; EQS Integrity Line, available at: https://www.eqs.com/en
-us/compliance-solutions/integrity-line/#features.
69 For example, SecureDrop, available at: https:// securedrop .org/
; Digital Whistleblowing Fund, availa-
ble at: https://www.hermescenter.org/supporting-diverse-initiatives-in-europe-through-the-use-of-secure
-whistleblowing-platforms/. For a comprehensive review of whistleblowing digital platforms used by jour-
nalists see Philip di Salvo, Digital Whistleblowing Platforms in Journalism. Encrypting Leaks (2020) Palgrave
Macmillan, at 63–89.
70 For example, SportsLeaks, available at: https:// www .sportsleaks
.com/; World Anti-doping Agency’s
SpeakUp, available at: https://speakup.wada-ama.org/WebPages/Public/FrontPages/Default.aspx;
International Olympic Committee’s Hotline, available at: https://ioc.integrityline.org/. See generally, United
Nations Office on Drugs and Crime, Reporting Mechanisms in Sport: A Practical Guide for Development
and Implementation (2019), available at: https://stillmedab.olympic.org/media/Document%20Library/
OlympicOrg/IOC/What-We-Do/Protecting-Clean-Athletes/Competition-manipulation/IOC-UNODC
-Reporting-Mechanisms-in-Sport-ebook.pdf.
71 For example, SEC Whistleblowing program, available at: https://www.sec.gov/whistleblower.
72 For example, NixWhistle proposed to create ‘CoronaSpeak’, encouraging people to report positive
COVID-19 cases, available at: https://www.nixwhistle.com/.
73 Social media has become a popular channel for online whistleblowing. Analysis of this phenomenon is
outside the scope of the present chapter; however, for some analysis of the role of social media for whistle-
blowers see: H. Latan, C.J. Chiappetta Jabbour, and A.B. Lopes de Sousa Jabbour, Social Media as a Form
of Virtual Whistleblowing: Empirical Evidence for Elements of the Diamond Model. J Bus Ethics (2020),
available at: https://doi.org/10.1007/s10551-020-04598-y
378
The oldest and most widely used technical applications are hotline services.74 15.030
They offer anonymity and increased accessibility, but at the same time, it is
impossible to share documents, expensive to maintain qualified operators,
who are able to work across different languages and time zones, and hard to
establish further contact unless the whistleblower calls again.75
The perceived effectiveness of web portals has prompted the emergence 15.033
of companies offering open-source whistleblowing software, providing any
organisation with tools to create their own whistleblowing web portal. Two
74 An interesting finding in one of the reported surveys is that the effectiveness of this technology increases
when it is branded as a ‘helpline’, rather than a ‘hotline’. See Stephen R. Stubben, Evidence on the Use and
Efficacy of Internal Whistleblowing Systems, 58 Journal of Accounting Research (2020), available at: https://
onlinelibrary.wiley.com/doi/10.1111/1475-679X.12303.
75 John Wilson, Whistleblowing: What are the Most Effective Speak-up Channels? (2017), available at: http://
in-houseblog.practicallaw.com/whistleblowing-what-are-the-most-effective-speak-up-channels/. For more
analysis of the hotlines’ effectiveness, see e.g., Eugene Soltes, Paper Versus Practice: A Field Investigation
of Integrity Hotlines, 58 Journal of Accounting Research (2020), available at: https://onlinelibrary.wiley.com/
doi/abs/10.1111/1475-679X.12302; Stephen R. Stubben, Evidence on the Use and Efficacy of Internal
Whistleblowing Systems, 58 Journal of Accounting Research (2020), available at: https://onlinelibrary.wiley
.com/doi/10.1111/1475-679X.12303.
76 Mostafa Hussien and Toshiyuki Yamanaka, Whistleblowing at Work. Can ICT Encourage Whistleblowing?,
27 Joho Chishiki Gakkaishi 150 (2017), available at: https://www.jstage.jst.go.jp/article/jsik/27/2/27_2017
_017/_article/-char/en, at 151.
77 Business Keeper AG, available at: https://www.business-keeper.com/en/whistleblowing-system/references
.html.
379
15.035 Wikileaks is another web portal, infamous for its major role in publishing mil-
lions of leaked documents, including the Iraq War Logs and Hillary Clinton’s
emails. Wikileaks has garnered significant controversy and the division
between public interest whistleblowing and politically-motivated leaking is
contested – an increasingly blurred distinction considered further below.
15.036 Such web portals have been required to adopt certain technological measures
to guarantee the security and anonymity necessary for whistleblowing. The
majority employ Tor, a ‘group of volunteer-operated servers’ that constitute
a distributed anonymous network. Servers of this network are connected via
virtual tunnels, concealing the path of a user’s traffic, ensuring privacy and
preventing tracking. Whistleblowing portals reliant on Tor adopt its ‘onion
78 For a further review of the features of GlobaLeaks and SecureDrop see Matthew Jenkins, Overview of
whistleblowing software, U4 Helpdesk Answer (2020), available at: https://www.u4.no/publications/overview
-of-whistleblowing-software.pdf.
79 GlobaLeaks, available at: https://www.globaleaks.org/about-us#our-vision.
80 Europe will begin to protect whistleblowers; institutions and firms must prepare (03 March 2020), avail-
able at: https://eatproject.eu/europe-will-begin-to-protect-whistleblowers-institutions-and-firms-must-pre
pare/.
81 See Expanding Anonymous Tipping website: https://eatproject.eu/.
82 SecureDrop, available at: https://securedrop.org/.
83 Amy Davidson Sorkin, Introducing StrongBox (2013), available at: https://www.newyorker.com/news/amy
-davidson/introducing-strongbox.
380
service’ in order to publish a website without revealing its location. This service
operates by using random ‘rendezvous-points’ where a client can go to access
the website, using a public key and an onion address, without revealing their
identity.84
2. Mobile apps
The use of mobile apps has been especially prominent on governmental levels in 15.038
countries with less developed technological infrastructure. In September 2020,
the Zimbabwe Anti-Corruption Commission introduced an anti-corruption
whistleblowing app.87 The app is expected to assist in anti-corruption matters
by allowing citizens to report ‘safely, loudly and visibly’, submitting evidence
necessary to launch anti-corruption investigations. The added benefit, relevant
in times of COVID-19 pandemic, is the minimum social contact required for
reporting.88 Similar apps, intended to provide a secure and private outlet for
blowing the whistle, were launched in Nigeria (‘Wahala Dey’),89 in Abu Dhabi
(‘Inform the Prosecution’)90 and in India, exclusively for members of political
party Makkal Needhi Maiam in order to flag issues caused by party members
(‘Maiam Whistle’).91
381
3. Blockchain
a. Anonymity
15.040 Given the ease of tracking the identity of whistleblowers when communicat-
ing online,92 blockchain is a potential solution which can strike the balance
between the need for anonymity and the importance of an investigative author-
ity being able to contact the whistleblower for further details. A project called
WhistleAI is currently working towards realising this potential by combining
the benefits of blockchain, crowdsourcing and AI. To ensure anonymity their
platform relies on zero-knowledge protocols, which entails splitting informa-
tion into fragmented pieces before sending it to the nodes for verification. This
ascertains the protection of whistleblower identity while allowing the members
of the network to verify the correspondence of the whistleblower’s allegation
with the information provided in their report.93
b. Immutability
15.041 The second advantage of blockchain is its immutability. Data, once uploaded
on a blockchain-based platform, cannot be deleted or tampered with as it is
aggregated into interconnected blocks. This prevents employers or organisa-
tions implicated by the disclosure from concealing the whistleblowing report.
An additional function of this platform may be public time stamping, which
allows whistleblowers to aggregate data for a period of time, before deciding
whether to publish the materials or not.94 Time stamping and immutability
of data would mean information could be used in future court proceedings
without concern for the veracity of evidence.95
c. Resilience
15.042 Relying on blockchain for whistleblowing would drastically increase the resil-
ience of the whistleblowing platform. Unlike website-based platforms, block-
92 Owen Bowcott, Whistleblowers Endangered in Digital Age, Says Lawyers’ Report (2017), available at:
https://www.theguardian.com/media/2017/feb/22/.
93 WhistleAI, available at: https://www.whistleai.io/WhistleAI.pdf.
94 Shafi Goldwasser and Sunoo Park, Public Accountability vs. Secret Laws: Can They Coexist? A Cryptographic
Proposal (2017), available at: https://eprint.iacr.org/2018/664.pdf, at 4.
95 Wolfie Zhao, Chinese Supreme Court Admitted Blockchain Evidence as Legally Binding (2018), available
at: https://www.coindesk.com/chinas-supreme-court-recognizes-blockchain-evidence-as-legally-binding/.
382
d. Compensation
A unique and arguably controversial feature of blockchain is its ability to offer 15.043
compensation to the whistleblower through smart contracts. This mechanism
would not only offer the whistleblower confidence of their identity’s security
through blockchain, but also provide them with adequate compensation
through the use of cryptocurrency, which could be automatically transferred to
their account once the ‘leaked’ data is verified and the appropriate conditions
for reward are satisfied. This idea has been implemented in WhistleAI, where
a privacy coin named WISL is used both for compensating whistleblowers
and for incentivising crowdsourcing participants that allow the platform to
continue operating.97
e. Escrow
Another possible advantage of blockchain is its application as an information 15.044
escrow. This can be done through a smart contract, programmed to release
information only if certain conditions are satisfied. For example, Callisto,
initially designed to combat sexual harassment on college campuses, forwards
the reported misconduct only when there are at least two complaints about
the same perpetrator.98 Such technology, combined with the other benefits of
blockchain, would help eliminate the ‘first-mover disadvantage’ and lessen the
likelihood of retribution.99
f. Concerns
There are, of course, potential disadvantages to blockchain’s adoption in the 15.045
present context. First, employing a distributed network means that all users of
a blockchain could have the sensitive data on their nodes (computers), poten-
tially exposing them to liability in some jurisdictions.100 One way in which this
96 Yochai Benkler, A Free Irresponsible Press: Wikileaks and the Battle Over the Soul of the Networked Fourth
Estate, 46 Harv. C.R.-C.L. L. Rev. 311 (2011), available at: http://benkler.org/Benkler_Wikileaks_current
.pdf, at 3.
97 WhistleAI, supra note 93.
98 Callisto project, available at: https://www.projectcallisto.org/.
99 Ian Ayres and Cait Unkovic, Information Escrows, 111 Michigan Law Review 145 (2012), available at:
https://repository.law.umich.edu/mlr/vol111/iss2/1, at 3; and Carsten Tams, Can ‘Allegation Escrows’
Remedy the Underreporting of Sexual Harassment? (2017), available at: http://www.fcpablog.com/blog/
2017/11/20/carsten-tams-can-allegation-escrows-remedy-the-underreportin.html.
100 Roman Matzutt et al., A Quantitative Analysis of the Impact of Arbitrary Blockchain Content on Bitcoin,
(2018), available at: https://www.martinhenze.de/wp-content/papercite-data/pdf/mhh+18.pdf, at 7.
383
risk may be mitigated is through the use of Enigma secret contracts.101 These
smart contracts use ‘secure computation’ technologies to compute over the
encrypted data, thus concealing sensitive information contained in the report
from the other members of the network, but retaining their ability to validate
the transactions.102 Secret contracts, thereby, offer privacy to the whistleblow-
ers, hiding their identity and mitigating the risk of retaliation. A second poten-
tial drawback of blockchain is that there has to be an established mechanism
of incentives to continue mining and, consequently, verifying the information.
Employing cryptocurrencies for this purpose might be one of the solutions, for
example, as demonstrated by the WhistleAI project.
g. Application
15.046 Today there are very few blockchain-based whistleblowing platforms. One
of the only examples is NixWhistle, a whistleblowing platform, built on an
open-source blockchain technology ‘Corda’.103 It operates by assigning to
a user one of three roles: whistleblower, investigator and reviewer. Such clear
division of roles, coupled with the benefits of the blockchain technology, has
allowed NixWhistle to ensure the anonymity of the whistleblower, integrity of
the reported data at every stage of the process and adherence to the role-based
access to information.
384
Some 50 years ago, Daniel Ellsberg spent 18 months meticulously copying 15.049
page after page of incriminating materials to reveal the Pentagon Papers.105 It
took only one memory card and several clicks for Edward Snowden to expose
gigabytes of data to the masses.106 Indeed, the most notorious of recent whis-
tleblowing incidents have taken the form of massive dumps of information to
web-sources: Snowden’s 2013 disclosure was 60GB in size; Antoine Deltour’s
2014 Luxleaks were 4GB; while the 2016 Panama Papers included 2.6TB
of information.107 Moreover, all of them contained massive collections of
documents containing a multitude of revelations of which the whistleblowers
themselves may have been unaware.
By radically affecting the sheer amount of information that can be disclosed 15.050
and the means by which to do so, technology is altering the nature of contem-
porary whistleblowing by blurring the lines between ‘whistleblowers’, ‘hackers’
and ‘leakers’. The implications of these developments must be confronted if
whistleblower protections worldwide are to be fit for the future.
105 Daniel Ellsberg, Secrets: A Memoir of Vietnam and the Pentagon Papers (2003), at 301.
106 Richard J. Aldrich and Christopher R. Moran, ‘Delayed Disclosure’: National Security, Whistle-Blowers and
the Nature of Secrecy, Political Studies (2018), available at: https://journals.sagepub.com/doi/abs/10.1177/
0032321718764990, at 7.
107 Suelette Dreyfus, Chelsea Manning and the Rise of ‘big data’ Whistleblowing in the Digital Age (2018),
available at: https://theconversation.com/chelsea-manning-and-the-rise-of-big-data-whistleblowing-in-the
-digital-age-102479.
108 OECD, Committing to Effective Whistleblower Protection (2016), available at: https://www.oecd.org/daf/
anti-bribery/Committing-to-Effective-Whistleblower-Protection-Highlights.pdf, at 18.
109 Ashley Savage, Whistleblowers for Change: The Social and Economic Costs and Benefits of Leaking and
Whistleblowing (2018), available at: https://www.opensocietyfoundations.org/sites/default/files/20181120
-whistleblowers-for-change-report.pdf, at 7.
110 Ibid.
385
15.052 The third dimension complicating this emergent dynamic is the rise of hackers
fulfilling a whistleblower-like function. This is illustrated by the ‘rise of cyber-
security whistleblowers’,111 who have been thrust into the limelight following
incidents such as the WannaCry ransomware attack, the Equifax breach and
the Cambridge Analytica Facebook data breach. This trend has only been
underscored during the COVID-19 pandemic. Cybersecurity whistleblowers
can include a range of actors from non-technical company employees uninten-
tionally made aware of security flaws or unreported data breaches, to ‘ethical’
or ‘white-hat’ hackers, cybersecurity professionals or hobbyists who conduct
solicited or unsolicited hacks and disclose any discovered vulnerabilities to
the public (‘full disclosure’) or to the company (‘coordinated/responsible
disclosure’), without exploiting those flaws.112 For example, since revelations
about Russian interference in the 2016 US election emerged, a group of ethical
hackers has turned their attention to election security, even spending their
own money to buy electronic voting machines for study.113 Many companies
are increasingly investing in the services of ethical hackers, as evidenced not
only by the demand for employees and contractors, but also by the growth of
‘hacking as an industry’, which includes crowdsourced cybersecurity and the
burgeoning freelance ‘bug-hunter’ bounty market, under which companies are
amending their terms of service to include standardised safe harbour provisions
for good-faith security research114 and incentivising ethical hackers to hunt for
vulnerabilities by paying for discovery.115
111 Dallas Hammer and Evan Bundschuh, The Rise of Cybersecurity Whistleblowing (2016), available at:
https://wp.nyu.edu/compliance_enforcement/2016/12/29/the-rise-of-cybersecurity-whistleblowing/.
112 National Cyber Security Centre, Coordinated Vulnerability Disclosure: The Guideline (2018), available at:
https://www.ncsc.nl/english/current-topics/responsible-disclosure-guideline.html.
113 Chris O’Brien, How ethical hackers are trying to protect the 2020 U.S. elections (2020), available at: https://
venturebeat.com/2020/10/23/how-ethical-hackers-protect-2020-u-s-elections/.
114 Disclose.io, Safe, Simple, Standardized Vulnerability Disclosure (2020) available at: https://disclose.io/.
115 Bugcrowd, A New Decade in Crowdsourced Security (2020) available at: https://www.bugcrowd.com/blog/
3-major-security-priorities-in-the-covid-19-era/.
116 Ibid.
386
1. Political implications
117 At the time of writing, the outcome of the case is unknown. Amit Yoran, The Future Of Cybersecurity Law
Hinges On The Supreme Court, available at: https://www.forbes.com/sites/amityoran1/2020/11/16/the
-future-of-cybersecurity-law-hinges-on-the-supreme-court/?sh=6ee680c8528a; for commentary and case
updates, visit: https://www.scotusblog.com/case-files/cases/van-buren-v-united-states/.
118 University of South Florida, Research Shows a 715% Increase in Ransomware Attacks in 2020 (2020),
available at: https://cyberflorida.org/covid/bitfender/.
119 World Health Organization, WHO reports fivefold increase in cyber attacks, urges vigilance (2020), avail-
able at: https://www.who.int/news/item/23-04-2020-who-reports-fivefold-increase-in-cyber-attacks-urges
-vigilance.
120 Madrea Matwyshyn, Hacking Speech: Informational Speech and the First Amendment, 107 North Western
U. L. Rev. 795 (2013), available at: https://scholarlycommons.law.northwestern.edu/nulr/vol107/iss2/10/.
121 Will Ma et al., Whistleblower or Leaker? Examining the Portrayal and Characterization of Edward Snowden
in USA, UK, and HK Posts, in Ma et al. (eds) New Media, Knowledge Practices and Multiliteracies (2014),
available at: https://www.researchgate.net/publication/265121066_Whistleblower_or_Leaker_Examining
_the_Portrayal_and_Characterization_of_Edward_Snowden_in_USA_UK_and_HK_Posts.
387
2. Ethical implications
15.057 The lack of responsible practices for publishing data leaks has already had con-
sequences. For example, in 2016 Wikileaks revealed 300,000 e-mails dubbed
the ‘Erdogan e-mails’. While subsequent investigation of these e-mails did
not yield significant evidence of wrongdoing, sensitive personal information,
including current phone numbers, citizenship IDs, addresses and political
party affiliations of millions of women, was released.125 Such data dumps have
122 Mary-Rose Papandrea, Leaker Traitor Whistleblower Spy: National Security Leaks and the First
Amendment, 94 B.U. L. Rev 449 (2014), available at: https://www.bu.edu/bulawreview/files/2014/05/
PAPANDREA.pdf, at 451.
123 John Wilson, Whistleblowing isn't Dobbing. It Supports our Democracy (2018), available at: https://www
.smh.com.au/public-service/whistleblowing-isnt-dobbing-it-supports-our-democracy-20180203-h0t6cv
.html.
124 Shkabatur, supra note 13, at 116.
125 Zeynep Tufekci, WikiLeaks Put Women in Turkey in Danger, for No Reason, The Huffington Post (2017),
available at: https://www.huffpost.com/entry/wikileaks-erdoganemails_b_11158792.
388
also revealed the personal information of rape victims and even the identities of
several gay men in Saudi Arabia, where homosexuality is illegal.126
These dilemmas demand a balancing test and – perhaps more than ever 15.058
before – there are no easy answers. On the one hand, excessive whistleblower
self-censorship is undesirable, in particular from a transparency perspective.
On the other hand, there is a need to minimise harm to those who, in the
absence of basic responsible data practices, may be unnecessarily compromised
as a side-effect of holding the powerful to account.127
3. Legal implications
Finally, changes in the nature and methods of whistleblowing are bound to 15.059
affect the drafting of new whistleblower protection laws that are being adopted
across the world, as well as the assessment of the extent to which existing
whistleblowing laws are fit for purpose. While there are many legal issues that
might arise, only a few relating to some of the key elements of whistleblower
protection schemes will be highlighted below.
First, as a result of these changes, the whistleblower, the regulator, any 15.060
whistleblowing service provider, and the entity accused of misconduct will
commonly be based in different jurisdictions. This puts pressure on jurisdic-
tional inconsistencies between the kinds of disclosure which will attract the
protection of whistleblower legislation. There is a range of sources of law
where whistleblower protections may be found, including bespoke legislation,
sectoral laws and laws specifically aimed at the public service. This can result
in legal loopholes which may deter potential cybersecurity whistleblowers who
are unsure whether they would be protected. Other laws take a more expansive
approach, capturing for instance disclosures in the ‘public interest’ or those
which disclose ‘abuse of laws’.128
For example, the law in the US is unclear about cybersecurity whistleblowing, 15.061
as there is no federal statute that directly addresses it. Instead, protection must
126 Nicky Woolf, WikiLeaks Posted Medical Files of Rape Victims and Children, Investigation Finds
(2016), available at: https://www.theguardian.com/media/2016/aug/23/wikileaks-posts-sensitive-medical
-information-saudi-arabia.
127 Alix Dunn, Responsible Data Leaks and Whistleblowing (2016), available at: https://www.theengineroom
.org/ r esponsible - data - leaks - and - whistleblowing/ ? fbclid = I wAR0 _ 8kIp L nGmSaCRZH C hpbiegf0d
WKd1XqXNdm7gSGGKdrIak1Vjc4sNUk.
128 OECD, G20: Study of Whistleblower Protection Frameworks, Compendium of Best Practices and Guiding
Principles for Legislation (2012), available at: https://star.worldbank.org/document/study-whistleblower
-protection-frameworks-compendium-best-practices-and-guiding-principles, at 6.
389
be read down from various existing federal or state laws.129 The SEC has been
taking a proactive approach to whistleblowing, and cybersecurity whistleblow-
ing in particular. In 2011, the SEC’s Division of Corporate Finance called for
the disclosure of cybersecurity incidences materially relevant to a company’s
operations as a part of regular reporting requirements under the federal secu-
rities regulation.130 In 2018, the SEC reiterated this call and offered further
interpretive guidance.131 However, this is not legally binding, giving rise to
a ‘grey area’ with respect to whether cybersecurity whistleblowers can take
advantage of the robust protection under the SEC’s whistleblower protection
programmes and the Dodd-Frank Act.132 Meanwhile, potential cybersecurity
whistleblowers who work on entities not regulated by these federal statutes are
left to fend off potential criminal liability for their actions.133 Even if Van Buren
is decided in favour of petitioners, for years security researchers have reported
a chilling effect on their work, admitting that ‘facing legal action is just one
of those things where it’s just not worth it anymore’.134 By contrast, Article
1 of the European Union Directive on Whistleblowing ‘lays down common
minimum standards for the protection of persons reporting (on) unlawful
activities or abuse of law’ and specifically includes ‘protection of privacy and
personal data, and security of network and information systems’.135 While
the inclusion of cybersecurity whistleblowers is a positive development, there
remains a lack of legal clarity over intersection with criminal laws directed at
not dissimilar conduct.136
15.062 Some question whether even broader whistleblower laws should capture the
reporting of ethical or immoral conduct, especially where ‘these tread the
fine line between illegality and morality’.137 On the one hand, it is argued
that extending protections to the disclosure of ethical or moral concerns may
129 Alexis Ronicker, Cybersecurity Whistleblower Protections (2017) available at: https://www.kmblegal.com/
sites/default/files/cybersecurity-whistleblower-protection-guide.pdf.
130 Jennifer M. Pacella, The Cybersecurity Threat: Compliance and the Role of Whistleblowers, 11 Brook. J.
Corp. Fin. & Com L. (2016), available at: https://brooklynworks.brooklaw.edu/bjcfcl/vol11/iss1/3/ at 50.
131 Securities and Exchange Commission, Statement and Guidance on Public Company Cybersecurity
Disclosures (2018) available at: https://www.sec.gov/rules/interp/2018/33-10459.pdf.
132 Pacella, supra note 130.
133 Ronicker, supra note 129.
134 Zach Whittaker, Lawsuits Threaten Infosec Research—Just When we Need it Most (2018), available at:
https://www.zdnet.com/article/chilling-effect-lawsuits-threaten-security-research-need-it-most/.
135 Article 2(1)(a)(x) Directive (EU) 2019/1937 of the European Parliament and of the Council of 23 October
2019 on the protection of persons who report breaches of Union law.
136 Centre for European Policy Studies, Software Vulnerability Disclosure in Europe: Technology, Policies and
Legal Challenges (2018), available at: https://www.ceps.eu/publications/software-vulnerability-disclosure
-europe-technology-policies-and-legal-challenges, at 42.
137 Savage, supra note 109, at 16.
390
This gap and the question of whether disclosures of ‘immoral’ conduct should 15.063
entitle whistleblowers to protection may, for instance, be directly relevant to the
joint initiative launched by a whistleblower non-profit ‘the Signals Network’ in
collaboration with international media groups. With a combined audience of
46 million people, this consortium is encouraging whistleblowers who believe
that corporations are ‘misusing’ big data to come forward.140 Consider the
GDPR, which took six years to come into force – although it has been hailed
as setting the global standard, some argue that its primary focus on individual
privacy rights and the protection of personally identifiable data is already
outdated. This on the basis that it fails to ‘account for the actual technological
landscape unfolding before us’, where the scale of big data analysis is such that
many of the most powerful applications and risks of harm are directed not at
individuals, but at groups.141 Imagine a data scientist at a start-up who, encour-
aged by the Signals Network Initiative, came forward disclosing that his/her
company was engaged in what he viewed as the use of big data in a way which
posed harm to a group.142 Given the relative underdevelopment of law relating
to group privacy harm, it is unlikely to be clear whether the practice disclosed
is unlawful or ‘merely’ immoral, and therefore whether whistleblower protec-
tions apply. As the law struggles to keep up with technological development
and the ethical issues it raises, there will be increased pressure on developing
whistleblowing legislation to adopt an expansive approach to the breadth of
protection, while intensifying the debate over whether to extend the scope of
protection schemes to the reporting of immoral or unethical conduct.
391
15.065 Notions of good faith and the relevance of motive are also inherent in the
concept of a whistleblower-as-hacker. Countries where the public prosecutor
can exercise discretion in pursuing cases throw this into sharp relief. Article 2
of the EU’s Cybercrime Directive and Article 3 of the Cybercrime Convention
lay down provisions regarding illegal access to information systems. However,
the notion of ‘ethical hacking’ does not exist in the criminal law. In order to
decide whether or not prosecution would be in the public interest, prosecutors
are relying on assessments of security researchers’ ‘bona fides’ and motives
to distinguish ‘white-hat’ and ‘black-hat’ hackers.148 It could be argued that
the distinction should instead be drawn by examining the proportionality of
the hacker’s actions: whether they did more than was necessary to expose
the breach. However, this determination is not free from difficulty, and for
an ethical hacker this may be impossible to predict in advance. This suggests
that an examination of motives and good faith will, in the context of new-age
whistleblowers, continue to be relevant.
392
The lines between whistleblowing, leaking and hacking are becoming blurred 15.067
in the age of information, changing the nature and methods of whistleblow-
ing. Protection of the whistleblowers of the future will depend not only on
addressing the political and ethical implications of these developments, but
also a worldwide effort to confront the need for protections responsive to the
globalised nature of modern whistleblowing.
E. CONCLUSION
Whistleblowing has changed considerably since the ancient Athenians hailed 15.068
the important function undertaken by those who drew public attention to
private wrongdoing. The most dramatic developments have occurred in the
149 Ashley Savage, Embracing the Challenges and the Opportunities of Cross-jurisdictional Whistleblowing
(2018), available at: http://www.oecd.org/corruption/integrity-forum/academic-papers/Savage.pdf.
150 Shu-Yi Oei and Diane M. Ring, Leak-driven Law, 65 UCLA Law Review (2018), available at: https://papers
.ssrn.com/sol3/papers.cfm?abstract_id=2918550, at 14.
151 Ashley Savage and Richard Hyde, Whistleblowing Without Borders: The Risks and Rewards of Transnational
Whistleblowing Networks, in David Lewis and Wim Vandekerckhove (eds) Developments in Whistleblowing
Research (2015), available at: http://www.track.unodc.org/Academia/Documents/151110 IWRN ebook
2015.pdf.
393
394