The UK’s financial regulator plans to use more AI in its own supervision as it responds to risks that Nikhil Rathi says are growing faster than traditional oversight can follow. The FCA chief also says it will write new rules where needed, rather than rely on rule-making alone.
Watch Desk analysis
What happened
Rathi made the remarks at the City Dinner at Mansion House on 22 October, according to Resultsense. He warned that AI and other fast-moving technologies are amplifying market risks, including volatility, concentration, market abuse and cross-border financial crime.
The FCA is already using AI tools to help supervisors build a fuller picture of risk. Rathi says agents can review cases, summarise returns and scan large datasets for gaps, freeing staff to assess harder questions, such as whether coordination between agents amounts to market manipulation. He also flagged the possibility that AI-assisted legal claims could add pressure to an already strained judicial system.
Why it matters
This is more than a warning about what AI might do to financial markets. Rathi describes a regulator trying to apply the technology to the data and cases it oversees, while keeping human judgement for decisions that do not fit neatly into an automated scan.
For firms, the practical signal is that existing reporting could face closer scrutiny as the FCA increases its use of AI. Rathi says new rules will be introduced where needed, but argues that regulation alone risks “chasing yesterday’s market”. His comments describe an approach, not a newly announced rule or a detailed timetable.
Our read
The interesting shift is that AI is part of both sides of the regulatory problem: agents may create new forms of market abuse, while the FCA wants to use AI to spot risks sooner. That could make supervision more responsive, but the difficult calls still need accountable people, not just a brighter searchlight over the same data.
What to watch
- Whether the FCA sets out how AI-assisted supervision affects firms’ reporting and examinations.
- Any new rules the regulator proposes in response to emerging AI-related risks.
- How the FCA assesses possible coordination between agents as market manipulation.
- Whether the regulator explains how supervisors check the quality of AI-generated summaries and alerts.
Discussion spark: Should financial regulators use AI to scrutinise firms’ data more aggressively, or could that create new risks if the system’s alerts are difficult to challenge?
Sources and evidence
- FCA chief: AI puts market integrity risks 'on steroids' – Resultsense (26 October 2026, 09:06 UTC)
- Telegraph job cuts as Axel Springer bets on AI publishing – Resultsense (26 October 2026, 09:06 UTC)
- OpenAI cuts frontier training off from the live internet – Resultsense (26 October 2026, 09:06 UTC)
- AISI: Mythos 5 misjudged real targets as simulated – Resultsense (26 October 2026, 09:06 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.