How the Bank of England, Federal Reserve, and FSB Are Regulating Autonomous AI Trading Agents

Central bankers in the UK, US, and globally are racing to understand and contain the financial stability risks posed by increasingly autonomous AI trading agents. The Bank of England has signalled a need for bespoke regulation, the Federal Reserve's new chair has called AI a "monumental paradigm shift," and the Financial Stability Board has published a set of sound practices for the financial sector. This article fact-checks the key regulatory measures and warnings from each institution, drawing exclusively on published sources from 2025 and 2026.

Bank of England: From Technology-Neutral to Bespoke AI Rules

The Bank of England (BoE) is actively moving toward a new regulatory framework for autonomous ("agentic") AI systems — those capable of making independent decisions and executing actions without human intervention.

Bespoke agentic AI rules. On June 30, 2026, BoE Deputy Governor Sarah Breeden signalled that agentic AI may require new regulatory reforms, marking a departure from the BoE's long-standing technology-neutral stance. Breeden said the BoE is considering whether banks need specific new rules for agentic AI and whether existing capital requirements adequately cover the risks .

AI stress testing and scenario analysis. The BoE confirmed in April 2026 that it is running scenario analyses and simulations to assess how AI trading agents could trigger financial instability, with a particular focus on correlated "herding" behavior — the risk that multiple AI agents acting on similar signals could amplify market swings . The UK Treasury Committee had recommended AI-specific stress tests in January 2026 .

Prudential Regulatory Authority (PRA) engagement. In its June 2026 joint response to the Treasury Committee, the BoE and PRA said they are promoting "responsible adoption" while keeping existing regulation "flexible, outcomes-focused, and effective" — but acknowledged the pace of AI evolution demands ongoing adaptation .

Broader financial stability monitoring. The BoE's December 2025 Financial Stability Report warned that deeper links between AI firms and credit markets, plus rising interconnections, mean an asset price correction in AI-linked lending could amplify stability risks . An April 2025 "Financial Stability in Focus" report flagged risks from AI use in banks' core decision-making, including potential credit risk spillovers .

Federal Reserve Chair Kevin Warsh: AI as a 'Monumental Paradigm Shift'

Kevin Warsh, confirmed as Federal Reserve Chair in 2026, has framed AI as both a transformative economic force and a source of potential disruption.

At the ECB Forum in Sintra, Portugal on July 1–2, 2026, Warsh called the rapid growth of AI a "monumental paradigm shift" for policy operations and the broader economy, adding that this revolution is "still in its first or second inning" . Central bank governors at the forum echoed concerns about AI-driven economic shocks .

Warsh has consistently argued that AI will act as a "significant disinflationary force" through productivity gains, which he has said could pave the way for lower interest rates . However, critics warn his view that AI is a guaranteed disinflationary force could risk premature rate cuts .

It is important to note that the available sources do not show Warsh issuing explicit warnings about AI disrupting labor markets or bank lending specifically — his public remarks are focused more on AI's macroeconomic and disinflationary potential. That gap is acknowledged here as a limitation of the available evidence.

Financial Stability Board (FSB): 12 Sound Practices for Responsible AI

The FSB moved from monitoring to active rule-making in June 2026. On June 10, 2026, the FSB published a consultation report on "Sound Practices for Responsible Adoption of Artificial Intelligence," setting out 12 sound practices for financial institutions covering governance and the full AI lifecycle .

The FSB "strongly" recommended that authorities and financial institutions impose tighter controls on agentic AI, warning that increasingly autonomous systems heighten risks for the financial sector as adoption accelerates . The framework is principles-based and non-binding, but signals a coordinated global push for governance standards around AI in finance .

Broader Central Banking Context

• European Central Bank (ECB): Research found that Q-learning (reinforcement learning) trading algorithms achieved high coordination but were prone to extreme bank run-like dynamics, while large language models produced heterogeneous and unpredictable behavior .
• European Systemic Risk Board (ESRB): Concluded in December 2025 that failing to keep pace with AI in finance could lead to increased financial instability and more frequent interventions by authorities, including risks from AI-driven changes in labour markets .
• Bank of England (December 2025 Financial Stability Report): Highlighted that AI transitions must occur "without compromising the resilience of the financial system" .

Key takeaway: The Bank of England is moving toward bespoke rules and AI-specific stress tests for autonomous trading agents. Kevin Warsh calls AI a "monumental paradigm shift" still in early innings and a likely disinflationary force. The FSB is consulting on 12 sound practices for responsible AI adoption, strongly urging tighter controls on agentic systems. Multiple central banks warn that unsupervised AI trading could trigger herding, bank-run-like dynamics, and broader financial contagion.