AI Due Diligence Agent Lands as EDD Enforcement Fines Climb

Sigma360 has launched an AI due diligence agent that automates the investigative legwork behind enhanced due diligence (EDD), the high-risk customer checks that have become a repeat target of regulatory enforcement. The New York risk-intelligence firm says the agent can cut case-handling time by up to 87% by taking over data gathering, evidence synthesis and the drafting of a cited investigation narrative, while leaving the human analyst to make and own the final risk decision. Announced on 22 October 2026, the tool arrives as the cost of getting EDD wrong is rising sharply.

The economics Sigma360 is attacking are specific. The company puts the average EDD case at four hours of analyst time, with roughly 85% of that spent on data gathering alone. At a mid-sized firm running thousands of cases a year, it estimates that adds up to around $2m in analyst capacity annually. Founded in 2016 and led by former US Treasury and National Counterterrorism Center official Stuart Jones Jr, Sigma360 raised a reported $17.3m in March 2026 and was ranked the top adverse media screening platform by Chartis Research in 2025 and 2026.

What does Sigma360's EDD agent actually do?

It automates the steps that drain the analyst's day, then hands the judgement back. The agent collects internal and external data, summarises findings across configurable modules covering industry risk, source of wealth and source of funds, beneficial ownership, geographic risk and others, flags risk factors and mitigants, recommends next steps and produces a fully cited narrative formatted for regulatory review. Sigma360 says the analyst stays the decision-maker throughout, with every risk rating and action logged and traceable at examination, and that customers can inspect the agent's underlying prompt to see what it does at each step.

The data layer is the part competitors will find hardest to copy. Rather than searching the open web, the agent investigates against Sigma360's proprietary risk dataset: sanctions and watchlist records, corporate registry and beneficial ownership data across more than 230 countries, and curated intelligence on financial crime networks, built over years of running the company's screening platform. The EDD agent sits inside that wider platform alongside sanctions and watchlist screening, adverse media monitoring, entity resolution, perpetual KYC and network risk detection.

Two numbers anchor the pitch, and both carry a caveat. Sigma360 claims up to 87% lower case-handling time and up to 90% lower EDD cost per case. Those are vendor projections rather than independently audited results, and buyers will want to test them against their own caseloads before budgeting around them.

Why are regulators making this urgent now?

Because supervisors have started naming due diligence failures explicitly, and the penalties are large. Sigma360 says enforcement actions by the FCA and FinCEN citing EDD failings have produced roughly $160m in combined fines since mid-2025. The direction of travel is independently documented. The FCA fined Nationwide £44m in December 2025 for weaknesses in customer due diligence and monitoring, and Barclays around £42m for failings in handling a corporate client linked to money-laundering risk; Monzo was fined £21m in July 2025 for inadequate controls during a period of rapid growth. FCA fines tied to anti-money laundering failures topped £124m across 2025. In the US, FinCEN imposed an $80m penalty on a broker-dealer in March 2026, its largest Bank Secrecy Act action against a broker-dealer, citing due diligence and surveillance gaps.

The common thread in those cases is that regulators are now judging demonstrable effectiveness over well-written policy. That is the pressure Sigma360 is selling into: a consistent, fully documented EDD file for every case, regardless of analyst seniority or caseload.

How does this fit the wider shift to agentic AI in compliance?

Sigma360 is moving into a lane the rest of the regtech market is already racing for. Napier AI's 2026 outlook names agentic AI as one of the defining trends in financial crime compliance for the year, describing AI as having crossed from experiment to expectation by the end of 2025. WorkFusion, now part of UiPath, argues that regulators no longer merely tolerate AI agents in AML operations but increasingly expect them, with early adopters deploying agents across sanctions-alert review, transaction-monitoring investigations and EDD. Rivals building in the same territory include ComplyAdvantage, alongside a crowded field of screening and investigation vendors.

That crowding reframes the competitive question. The contest is no longer whether AI reaches EDD, but which vendors can prove their output is explainable and audit-ready under a supervisor's eye.

What is the catch for compliance teams?

The hard part is defensibility, not speed. An agent that drafts a due diligence narrative in minutes only helps if an examiner accepts that narrative as a sound, explainable basis for a decision. Napier AI's own forecast stresses that the 2026 priority for institutions should be explainability and guardrails over autonomy, precisely because supervisors are scrutinising whether AI acts in ways that cannot be audited. Sigma360's answer is to keep the analyst as the named decision-maker, expose the agent's prompt and cite every source behind each conclusion. Whether that framework satisfies an examiner reviewing an AI-assisted file is the question that will decide how fast banks actually adopt it.

Why This Matters to FinanceX Readers

For compliance leaders and the investors watching regtech, this launch is a test of a specific thesis: that enforcement pressure will push institutions to automate the most labour-intensive corner of financial crime compliance faster than they can hire for it. The firms that benefit are those that can show an examiner a fast, consistent and fully documented EDD file; the risk is buying speed that collapses under scrutiny. With the FCA and FinCEN both signalling that due diligence quality is now an enforcement priority, the budget case for agentic EDD is strengthening, and so is the standard any tool will be measured against.