FCA Plans AI-Driven AML Supervision of 60,000 Firms by 2028

The UK Financial Conduct Authority (FCA) is developing artificial intelligence tools to support anti-money laundering (AML) supervision of approximately 60,000 firms expected to come under its remit in 2028. The initiative forms part of a wider strategy to use automation and AI to identify risks earlier, analyse information more efficiently and direct supervisory resources towards areas requiring closer scrutiny.

The plans were outlined by Jessica Rusu, the FCA’s chief data, information and intelligence officer, in a speech delivered on 19 October 2026 at Bloomberg’s UK Policy Series. The speech described the regulator’s efforts to incorporate AI into its own supervisory operations while responding to the increasing use of autonomous technology across financial services.

The planned expansion of AML oversight represents a significant increase in the FCA’s supervisory responsibilities. The approximately 60,000 firms are expected to include entities in the legal and accountancy sectors, whose AML supervision is due to transfer to the FCA towards the end of 2028.

The regulator is developing tools intended to help process information across this large population and identify potential warning signs. AI could assist in connecting information from different sources, detecting unusual patterns and helping supervisors prioritise cases for further examination. The FCA has not yet published a final technical specification for the planned system, and the initiative remains under development.

The FCA is also deploying automation and developing an AI agent known as the market first responder to support oversight of approximately 9,000 smaller wholesale-market firms subject to transaction-reporting requirements. This is a separate initiative from the planned AML supervision of 60,000 firms. The objective is to automate aspects of routine oversight so that supervisory staff can concentrate on more complex matters.

The wider programme includes work on machine-readable regulatory rules and data collection interfaces. These tools are intended to make regulatory information easier for firms and supervisory systems to process, potentially improving the speed and consistency of information exchanges. The speech did not announce a new reporting format or a specific implementation deadline for individual firms.

Alongside its plans for AI-assisted supervision, the FCA published examples of good and poor practices observed through its engagement with financial services firms. The examples address how businesses assess the need for AI, establish governance arrangements, test systems and monitor their performance. The publication is intended to evolve as the technology develops rather than serve as a fixed set of standalone rules.

The FCA emphasised three central principles for the use of AI in financial services: protecting consumers, maintaining clear accountability and ensuring operational resilience. These considerations are particularly relevant as systems move beyond analysing information and begin making decisions or executing actions with limited human intervention.

The emergence of agentic AI is a central part of the regulator’s strategy. Unlike tools that primarily provide recommendations or summaries, agentic systems can carry out tasks such as searching for products, managing payments and interacting with other systems on behalf of users or organisations.

The FCA plans to work with financial institutions, other regulators and AI developers on a framework for trusted agentic finance. Areas under consideration include agent identity, delegated authority, accountability and the execution of financial transactions. The work is intended to address questions about who authorises an AI agent to act, the limits of its permissions and how responsibility is established when something goes wrong.

Operational resilience is another priority. The regulator warned that an AI system that performs safely during initial testing may behave differently when deployed at scale, operating under pressure or using an updated underlying model. Firms therefore need to consider ongoing assurance, monitoring and arrangements for responding to system failures.

The FCA also highlighted the importance of reporting significant AI failures and near misses within the applicable regulatory framework. Its speech did not introduce a universal new reporting deadline for all AI incidents, but stressed that existing incident-reporting expectations must be considered as financial institutions adopt more autonomous systems.

Financial crime prevention is among the areas where AI may assist firms and regulators. Applications include fraud detection, analysis of suspicious patterns and the processing of large volumes of information. However, these capabilities do not eliminate the need for effective governance, reliable data and human judgement in consequential decisions.

For AML compliance teams, the FCA’s plans indicate a future in which supervisory technology may be used more extensively to analyse information and identify potential weaknesses across regulated populations. Firms should continue to meet their existing obligations concerning customer due diligence, transaction monitoring, suspicious activity reporting and internal controls. The regulator has not announced a new AI-specific AML rule as part of this initiative.

The FCA’s approach remains based on its existing regulatory framework rather than the introduction of a separate set of AI-specific financial services regulations. The authority has stated that current requirements, including those relating to governance, accountability and consumer protection, are intended to apply as firms adopt new technologies.

The regulator’s wider AI programme follows the publication of the Mills Review in July 2026, which examined how artificial intelligence could reshape retail financial services by 2030 and beyond. The review considered potential benefits for consumers and businesses alongside risks involving fraud, cybersecurity, market competition and consumer harm.

The FCA’s October announcement sets out how it intends to apply AI within its own operations while continuing to monitor developments across the financial sector. Its planned AML supervision tools and separate wholesale-market automation initiative remain in development, with further work expected as the regulator prepares for its expanded responsibilities in 2028.

By fLEXI tEAM