We sat down with Mitch Trehan, Chief Compliance Officer at Allica Bank and a contributor to The Future of AI Governance and Compliance in Financial Services, following the roundtable discussion at its launch in the House of Lords. Mitch reflects on how the industry can collectively address AI governance, and what the compliance function of the future looks like.
1. Coordination across the industry is vital for sound AI governance
No settled operating model yet exists. The industry depends on collective learning through professional networks, trade bodies and the supervisory authorities.
2. AI will place a premium on the precision of policy documentation
Advisory work consists largely of answering first-line requests for guidance, much of which language models can retrieve from external regulation and internal policies, procedures and risk appetites. This places a premium on documentation quality. The more specifically policies are written, the more accurately systems can interpret them and support advisory queries.
3. Assurance will move from sample-based testing towards complete review
Assurance has relied on sample testing because of the time required to process samples. AI compresses that timeframe, allowing samples to be enlarged or extended to the full population. Mapped to defined MI thresholds, KPIs and KRIs, this improves both the confidence and completeness of assurance.
