The UK's Anti-Money Laundering and Asset Recovery Strategy 2026–2029, published in September, puts at least £550m and more than 500 new officers behind one idea: crime should not pay. Its message to regulated firms is just as clear: stop spending on compliance that doesn't catch criminals and be ready to prove that what you do spend works.
The strategy is unusually candid. Government admits that despite significant public and private investment, too much AML activity, particularly in the private sector, is still low-value "tick-box" compliance. Its first pillar, Target, commits to cutting that work and redirecting effort to the highest-harm risks in the 2025 National Risk Assessment.
We can already see which way this is going. A risk-based approach to enhanced due diligence on high-risk jurisdictions is expected to save firms £178.1m a year. The POCA threshold behind low-value DAML reports has risen from £1,000 to £3,000, and with 866,616 SARs filed in 2024–25, government will examine raising the suspicion threshold itself. Supervisors are to discourage blanket or excessive controls that disrupt legitimate customers.
That freedom comes with a condition. Over the next few years the number of AML supervisors will fall from 25 to 3, with the FCA taking over lawyers, accountants and trust and company service providers. HM Treasury and the supervisors will agree common principles for supervisory risk assessments — methodology, evidence base and update frequency — and Treasury will consult on stronger enforcement powers, including director accountability.
In practice, every decision to scale a control down will need a defensible rationale. That means a documented methodology, a clear link to the NRA and the published system priorities, and an audit trail of who decided what, when and why. A business-wide risk assessment kept in a spreadsheet and refreshed once a year will not carry that weight.
The second pillar, Integrate, creates a National Financial Intelligence Service in the NCA. It brings JMLIT and Data Fusion together, aligns the UK Financial Intelligence Unit, uses AI to spot patterns in large datasets, and extends beyond banks to crypto firms, money service businesses and e-money institutions — with a review of bringing in legal, accountancy, gambling and insurance.
The framing is national security as much as crime: hostile state actors, sanctions evasion and corrupt elites hiding assets behind opaque ownership. Meanwhile the FCA will publish good and poor practice on AI in AML during 2026–27. Regulators want AI adopted — but with proper governance. The line between a firm's risk assessment and the state's intelligence picture is getting thinner.
Acuminor was founded by people from law enforcement, military intelligence and banking to close exactly this gap. Our risk engine, fed by a structured intelligence stream drawn from 750+ sources on money laundering, terrorist financing, sanctions and proliferation financing, fraud and corruption, lets firms manage the whole financial crime framework on evidence: enterprise-wide risk assessment, threat assessment, customer risk scoring models, high-risk country models, framework optimisation, and transaction monitoring tuned to real typologies — cutting alert volumes by up to 90%. Every decision carries a documented rationale for auditors and supervisors. The same capability serves government and public-private partnerships. With AMLA setting a parallel course in the EU, firms on both sides of the Channel face the same test.
Three steps worth taking before the supervisory reforms take effect in 2028/29:
The strategy rewards firms that can show their thinking. The question is no longer whether you comply, but whether you can explain why your controls point where they do.