Can machine learning, as a RegTech compliance tool, lighten the regulatory burden for charitable organisations in the United Kingdom?

Published date06 August 2021
Date06 August 2021
Subject MatterAccounting & finance,Financial risk/company failure,Financial crime
AuthorCharanjit Singh,Lei Zhao,Wangwei Lin,Zhen Ye
Can machine learning, as a
RegTech compliance tool, lighten
the regulatory burden for
charitable organisations in the
United Kingdom?
Charanjit Singh
Holborn Chambers and Business School, University of Westminster, London, UK
Lei Zhao
School of Law, University of Chinese Academy of Social Sciences, Beijing, China
Wangwei Lin
Department of Law, Coventry University, Coventry, UK, and
Zhen Ye
3PB, London, UK
Purpose Machine learning is having a majorimpact on banking, law and other organisations. The speed
with which this technologyis developing to undertake tasks that are not only complex and technical but also
time-consuming and that are subject to constantly changing parameters is astounding. Thepurpose of this
paper is to explore the extent to which machinelearning can be used as a solution to lighten the compliance
and regulatory burden on charitable organisations in the UK; so thatthey can comply with their regulatory
duties and developa coherent and streamlined action plan in relation to technologicalinvestment.
Design/methodology/approach The subject is approached through the analysis of data, literature
and domestic and international regulation. The f‌irst part of the study summarises the extent of current
regulatory obligations faced by charities, these are then, in the second part, set against the potential
technologicalsolutions provided by machine learningas of July 2021.
Findings It is suggested thatcharities can use machine learning as a smart technologicalsolution to ease
the regulatoryburden they face in a growing and impactful sector.
Originality/value The work is original because it is the f‌irst to specif‌ically explore how machine
learningas a technologicaladvance can assist charities in meeting the regulatorycompliance challenge.
Keywords Machine learning, RegTech, English law, Unsupervised learning, CharityTech
Paper type Research paper
1. Introduction
Artif‌icial intelligence (AI) is changing the way in which organisations work. The ability of
AI to automate tasks that might be considered tediousgenerates immense benef‌its by
creating time for strategizingand networking. AI, data analytics and machine learning(ML)
have become commonplace buzzwords, their potential is being built into the fabric of
organisational technology systems as innovative solutions to issuesrelating to compliance.
In this article, we explore how ML and its constituents, i.e. unsupervised learning can, as a
Journalof Financial Crime
Vol.29 No. 1, 2022
pp. 45-61
© Emerald Publishing Limited
DOI 10.1108/JFC-06-2021-0131
The current issue and full text archive of this journal is available on Emerald Insight at:
RegTech and CharityTech tool, assist charities in meeting the regulatory compliance
challenge. In so doing we investigatewhether ML is a trustworthy component in the arsenal
of the not-for-prof‌it charitysector.
2. Practical and theoretical f‌inancial crime issues facing charities in the UK
In our previous article, we explored the f‌inancial crime issues facing charities (Singh et al.,
2020), both practically and theoreticallyand def‌ined the organisations that are the subject of
this research. Therefore,we do not propose to set that out again save in summary. Charities,
in England and Wales, are regulated organisationsthat are formed for particular charitable
purposes. In law, they are purpose trusts[1] without named benef‌iciaries. Although they fall
into the voluntary sector they are in fact distinguishableand the sector includes many non-
prof‌it and non-charitable organisationswhich can add to the complexity of issues discussed
in this research. Charities fall into the sector often referred to as the third sectorthat sits
alongside the publicand privatesectors.The legal def‌inition of charitable purpose and
the description of that purpose is set out in ss. 13 of the UKs Charities Act 2011. Save in
short, the latter is to prevent or relieve the poverty of the advancement of education or
religion etcetera[2].
The legal form that charities may take includes companies [3] limitedby guarantee with
trustees as board members. Of particular regulatory concern are shell charities, these are
shell corporations [4] or companies set up in compliance with the relevant legislation with
f‌inancial assets but they conduct littleor no business activity. The primary purpose of such
an organisation is to functionas a conduit for anonymous f‌inancial transactions.It is salient
to state that whilst they are used for legitimate purposes i.e. asset storage for start-ups, the
form is often exploited to further illegal purposes including money laundering. Shell
charities often fall into this lattercategory [5] and, therefore are a major f‌inancial crime risk.
The rules relating to charitiesmust be followed regardless of the legal form such
organisations take, the actions of decision makers are regulated by the rules of equity,
various f‌iduciary duties and the duty of prudence, care and skill as set out in the Trustees
Act 2000 [6]. It is the role of the Charity Commission to promote transparency in the
f‌inancial affairs of third sector organisations with the aim of sustaining and promoting
growth and donor trust in charitable giving. The f‌inancial i.e. tax benef‌its [7] to achieving
charitable status is a matter beyond the scope of this article but may present problems
relating to fraud in its own regard for the exchequer. In 2021, there were over 169,779
charities registered in England and Wales or 212,063 operating in the UK with 19,731
operating overseas as at 2018 per the FATF Mutual Evaluations report [8]. These f‌igures
pose a signif‌icant compliancechallenge both for the authorities and the charities themselves.
3. Machine learning and the regulatory risk and compliance function
Not-for-prof‌it fundraising is a human endeavourworthy of praise. The traditional methods
used to generate funds (Sheldon,2000) are; grants, networking and donor dinners and direct
giving from corporate foundations. The more run-of-the-mill meetings in coffee shops are
supplemented to fund the change-the-world initiatives. AI has proven benef‌its of achieving
low human resource costs whilst maintaining high levels of relationship-building and
outreach activity. In short, AI, of which ML is a constituenttechnology, allows a machine or
series of machines to act, comprehend, learn and sense just like humans would.
Unsupervised learning (UL),also a constituent or AI, is a form of ML in which the system is
trained to identify patterns in data sets where the right answermay not be apparent
because it is diff‌icult to determine, perhaps, due to the sheer quantity of data that needs
processing. UL can create outputs byclustering data together based on perceived patterns.

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT