REGULATION CHALLENGES

Updating the regulatory frameworks

“institutions must consider how to develop Al while staying compliant”

Interpreting regulatory requirements for Al-based solutions is complex, as these frameworks were not designed with Al applications in mind; this exposes institutions to making mistakes in the adoption of Al

Global and local data-sharing regulations

“Firms’ global ambitions will be determined by the regulatory environment of their local geographies”

The fragmented nature of regulation across nations segments the pace of Al innovation and the structure of financial services by region. Strategic national competition also makes regulatory coordination difficult, creating barriers for adoption and stalling investment

Lack of standards for personal data

“Increased flows of data are crucial to the enablement of Al, but what are the consequences?”

Realizing the benefits of Al requires the continuous flow of personal data across institutions. Robust authentication and verification methods that are standardized across institutions are necessary to ensure security and privacy, and minimize the risk of fraud or other undue harms

Complexity in auditing AI systems

“Neural network Al solutions are often complex black boxes”

While regulators require audits of new processes (e.g. to verify independence), many Al processes are not easy to audit (e.g. traditional interim working steps are often skipped entirely), raising questions around when an audit is or isn’t necessary

What is liability model?

“if Al-powered decisions result in amplified losses, who becomes liable?”

Al introduces ambiguity into responsibility in the event of loss or negative outcomes. Institutions are hesitant to use third-party services for Al systems as regulators will likely hold the primary institutions responsible if there are damages to recover

Tendency to favor risk mitigation

“New and unknown risks introduced by AI to firms and the public difficult to tackle”

Regulators are continuously challenged to create the optimal balance between enabling innovation and risks avoidance. However, the complexity surrounding Al, combined with regulators’ natural tendency to be cautious, suggests that regulatory frameworks are likely to lean in favour of risk mitigation