June 2026
KPMG recently hosted a joint webinar with the International Regulatory Strategy Group (IRSG), bringing together regulators and industry practitioners to explore how financial services firms are navigating the realities of AI adoption.
As artificial intelligence moves from experimentation into live deployment, firms are increasingly focused on:
Insights from IRSG’s paper on emerging global norms highlighted that AI adoption is accelerating globally. Regulatory approaches are broadly aligned on the principles of:
However, countries diverge in how they operationalise these principles from prescriptive regimes to principles-based and innovation-first approaches. Firms are often expected to use AI within existing regulatory frameworks.
Key challenge: operating and governing consistently across jurisdictions while meeting jurisdiction specific requirements.
The FCA perspective reinforced this principles‑led approach. Supervisory attention is on whether firms can demonstrate effective governance, good outcomes and ongoing assurance once AI systems are live.
Engagement with regulators, testing and controlled experimentation were highlighted as increasingly important as AI is deployed into more complex and higher‑impact use cases. The use of AI within regulatory authorities themselves further underlined that the technology is already operational, not theoretical.
Industry panellists discussed the practical realities of implementation. AI was described as amplifying familiar risks (both in speed and scale), particularly around accountability and explainability.
As models become more complex and data volumes grow, firms will need to continue to explain and defend AI‑driven decisions, especially in customer‑facing or higher‑risk contexts.
One clear area of change discussed was the evolving role of the three lines of defence. Governance and controls are moving decisively into the first line, with monitoring and explainability designed into AI systems from the outset. The second line is helping to define guardrails and risk indicators rather than relying on point‑in‑time validation.
Hear from Leanne Allen, UK Head of AI at KPMG in the UK talk about the shifting role of the three lines of defence in financial services firms as they embed AI.
As AI adoption scales and models develop, operational challenges include:
And therefore requiring:
Hear from Leanne Allen, Head of AI at KPMG in the UK, as she discusses the challenges of scaling AI and the risk posed by the use of legacy tech in a world of frontier AI models.
Beyond technical considerations, the panel highlighted that new forms of behavioural risk, such as complacency and ‘purpose creep’, emerge as AI tools are used more widely across organisations. Ensuring appropriate human oversight, escalation routes and designed‑in friction become increasingly important as AI moves from specialist teams into everyday operations.
Success will depend less on new regulation, and more on how effectively firms:
