Aim: To provide ways in which professionals and firms worldwide can respond to the ethical challenges of Artificial Intelligence (AI, machine-learning)
The Worshipful Company of Information Technologists, in association with the Chartered Institute for Securities & Investment (CISI), British Computer Society, BSI, United Kingdom Accreditation Service, London Chamber of Commerce & Industry, Northeastern University, and Z/Yen Group, with the cooperation of Citigroup, London Metal Exchange, and LearnerShape has launched an initiative to develop ethical AI standards. Initially the standards are for professionals working in AI. Later, the Initiative intends to look at standards for firmwide certification. The approach is international rather than national, and promotes the use of existing ISO standards rather than the creation of new ones. We have begun by creating an online course for professionals, www.cisi.org/ai.
AI has the potential to affect society and individuals materially in both positive and negative ways. Rapid development of AI technology means that ethical considerations must be taken into account from the beginning of the design process. AI systems rely on large amounts of data, which raise privacy concerns. Without ethical guidelines and regulations in place, the misuse or mishandling of data can result in harm to individuals or groups. Ideally, AI systems are transparent, accountable, and inclusive, and do not reinforce or amplify existing inequalities.
As AI becomes more advanced and widespread, there are growing concerns about issues such as bias, transparency, accountability, and safety. If people do not believe that AI is being developed and used in a responsible and ethical manner, they hesitate to use or interact with beneficial technologies.
On 27 June 2023 we launched the “695th Lord Mayor’s Ethical AI Initiative” on “Delivering Ethics Courses For AI Deployers & Builders” at CISI, starting in financial services.
This initiative aims to ensure that AI is developed and deployed in ways that benefit society, while minimising potential harm. Ethical AI training can help ensure that AI is developed and deployed in a way that aligns with our values, respects human rights, and promotes the common good. This initiative should establish ethical guidelines and principles for AI development, encourage transparency and accountability in AI systems, and promote collaboration and dialogue among stakeholders.
There are two obvious types of qualification:
1 – Domain qualifications for ‘AI Deployers’, e.g. the responsible deployment of AI. CISI would be the financial services domain certificate provider. Early discussions with other sectors indicate interest in areas such as law, accountancy, surveying, and pharmaceuticals. This takes the educational form of a Professional Assessment course, circa 12 to 20 hours. Refresher or update mini-courses might also be considered. CISI have volunteered to develop and administer this course for financial professionals.
2 – An ‘AI Builder’ qualification for software designers, engineers, and computer scientists. BCS would be the certificate provider. This is likely to take the form of a Professional Certificate, circa 100 to 150 hours. This is a significant undertaking, and BCS are observing the creation of the AI Deployer course before committing to an AI Builder course. This work is likely to be done in conjunction with the Open Data Institute who have made significant progress on a rich curriculum.
CISI is leading course development. The course is in six parts, with a starting outline of:
- establish endeavour – why? background to ethics and schools of ethics, particularly Principlism – https://en.wikipedia.org/wiki/Principlism, principles for AI
- UK government guidance – https://www.gov.uk/guidance/data-ethics-and-ai-guidance-landscape
- UK AI Regulation Approach (White Paper https://www.gov.uk/government/publications/ai-regulation-a-pro-innovation-approach & updated versions thereof)
- Universal Guidelines for Artificial Intelligence – https://archive.epic.org/international/AIGuidleinesDRAFT20180910.pdf
- G20 AI Principles – https://oecd.ai/en/wonk/documents/g20-ai-principles
- OECD AI Principles – https://www.oecd.org/digital/artificial-intelligence/
- Australia’s 8 Ethics Principles – https://www.industry.gov.au/publications/australias-artificial-intelligence-ethics-framework/australias-ai-ethics-principles
- assess & appraise – what? defining AI and finding AI
- ISO/IEC TR 24368:2022 standard and the relevant terms
- Proposal For A Regulation Of The European Parliament And Of The Council Laying Down Harmonised Rules On Artificial Intelligence (Artificial Intelligence Act) And Amending Certain Union Legislative Acts – https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=celex%3A52021PC0206
- lookaheads & likelihoods – where? what are the ethical issues and risks?
- “Information technology — Artificial intelligence — Overview of ethical and societal concerns”, PD ISO/IEC TR 24368:2022
- Information technology — Artificial intelligence — Guidance on risk management”, BS ISO/IEC 23894:2023
- Sources for keeping up-to-date
- options & outcomes – who? people and roles
- “Draft BS ISO/IEC 42001 Information Technology – Artificial intelligence – Management system”
- Understanding ‘stakeholders’
- understanding & undertaking – how?
- Z/Yen & BCS DIODE methodology has been contributed and used as an example – https://www.zyen.com/media/documents/ethical-assessment.pdf
- Citigroup methodology
- securing & scoring – when? six to ten case studies
In order to have a broad-based financial services perspective, a wider range of financial services firms are being sought for the first course, but there is already a sufficient number of prospective students (almost 100) to begin.
While the CISI course development is underway and led by CISI, the role of the Steering Group is just that, steering:
- making intellectual connections and contributions;
- making personal connections to people or organisations who might usefully contribute;
- providing a sounding board and critiquing function;
- publicising when appropriate, particularly towards the end of the Initiative.
WCIT and Professor Mainelli have appointed Liveryman Nicholas Beale as Chair of the Steering Group. The Steering Group is exploring how to provide regular updates for a rapidly changing field, including via CISI’s monthly ethics columns. We are actively exploring extending the course to other professions such as solicitors, barristers, surveyors, and accountants.
The next stage will be looking at how to certify organisations as adhering to Ethical AI.
For further information: