Ethics of AI 2024

In this IIL London Market webinar, Dr Artur Niemczewski discusses the Ethics of AI and in particular ethical considerations when using AI in insurance and the expectations of insurance professionals using it. Dr Niemczewski publishes a series of articles on AI within CII’s The Journal. 

Description

Despite popular beliefs, no AI model in use today has any in-built ethics. Ethics, i.e., ‘do no harm’ are only controlled through guardrails imposed by model curators. These are easily jailbroken by hackers and other bad actors. What can go wrong:

• Bias
• Hallucinations
• Malicious use

The UK Department for Science, Innovation and Technology published five principles underpinning responsible AI in the UK’s AI Policy Paper, ‘AI regulation: a pro-innovation approach”’ The EU has gone much further with the EU AI Act, which specifically lists prohibited uses. While there is no specific FCA regulation on the use of AI within financial services, the FCA expects that any outcomes are consistent with the existing regulation, in particular:

• Conduct Rules
• Consumer Duty

It is the responsibility of insurance professionals using and supervising AI to ensure ethical outcomes. The CII’s Digital Companion to the Code of Ethics provides an excellent reference.

Learning objectives

• Ethical considerations when using AI within the insurance value chain
• Regulators’ views
• ‘Human in the loop’ – expectations of insurance professionals using AI

Speakers

• Dr Artur Niemczewski, AI Champion, CII
• Chair: Caroline Wagstaff, CEO, LMG