Fighting AI with AI: The new battleground for investment firms

For investment firms, trust is the product. Clients hand over significant assets on the expectation that the firm verifying their identity, screening their counterparties, and monitoring their transactions can tell the real from the fake. Nowhere are the stakes higher than in enhanced due diligence (EDD) on high-net-worth individuals (HNWIs), politically exposed persons (PEPs), and the complex web of entities that surround them.

Generative AI (GenAI) is now directly testing that expectation from both sides. The same technology that lets firms onboard high-value clients faster is also giving criminals the tools to impersonate them, fabricate their documentation, and slip synthetic identities past controls built for an earlier era. 

This blog expands upon excerpts from Chapter 1 of our State of Financial Crime 2026 report, examining what that shift means for firms that compete on the quality of their due diligence.

A dual-use technology, weaponized against the investment sector

GenAI has reduced the barriers to entry for financial criminals to the point where it is a clear and present danger. The Financial Action Task Force (FATF), national regulators, and law enforcement agencies now treat GenAI-enabled crime as a mainstream threat, driven by a growing number of cases in which model-generated material has been used to conceal real identities.

Investment firms are squarely in scope. Scammers are using GenAI to generate content for clone investment websites that mimic the branding of well-known, regulated businesses or fabricate credible-looking new market entrants, complete with professional collateral and chatbots running coherent customer service scripts. 

The threat runs deeper than fake firms. For a firm onboarding an HNWI or politically exposed client remotely, GenAI makes it cheaper to fabricate a plausible source of wealth (SOW) narrative and generate a professional-looking backstory that a rushed manual review can wave through. Establishing a veneer of credibility is the entire game, and GenAI has made this cheap to produce at scale.

The sector’s exposure to illicit finance is also well understood by regulators. In the United States (US), the coming Investment Adviser AML Rule was introduced precisely because advisers manage vast pools of capital that corrupt officials, sanctioned actors, and foreign adversaries have sought to exploit. In the United Kingdom (UK) and across the European Union (EU), the Financial Conduct Authority (FCA) and evolving EU frameworks hold investment firms to stringent EDD standards for the same reason. GenAI raises the stakes on every one of these obligations.

In November 2024, the US Treasury’s Financial Crimes Enforcement Network (FinCEN) flagged a rise in suspicious activity reports (SARs), citing the use of high-quality fake IDs in fraudulent account openings. For firms whose entire risk posture rests on knowing exactly who a client is and where their money originated, this is the core problem, not a peripheral one.

Why “more analysts” is the wrong answer

The pattern most compliance teams recognize is that as GenAI-assisted deception gets harder to spot, individual cases take longer to clear, screening systems generate more alerts against obfuscated names and entities, and the review queue lengthens. Alert volumes climb, the EDD backlog grows, and complex SOW cases sit waiting. Facing that backlog, the instinctive response is to hire more analysts to work through the queue.

But the economics do not hold. Criminals using GenAI operate at a speed and volume that manual review cannot match, and they continuously refine their methods. In 2024, the Federal Bureau of Investigation (FBI) and the US Department of Homeland Security both reported that fraudsters are now A/B testing their scripts in real time.

A compliance function that meets industrialized deception with more headcount is bringing labor to an arms race. The cost compounds, the backlog persists, and the firm faces a familiar dilemma: onboard fewer clients and forgo growth, or revert to slow, costly, labor-intensive verification. Neither protects the firm, and both concede ground to competitors.

The more effective response is automation that matches the threat’s tempo and clears the backlog at its source. Firms are already deploying artificial intelligence (AI) to review customer due diligence (CDD) and know-your-customer (KYC) packs, reconcile data against registries, and generate risk profiles for expert review. 

According to our 2026 report’s global survey of 600 senior compliance professionals, 41% of organizations using, piloting, or evaluating advanced AI have implemented automated onboarding and KYC processes.

State of Financial Crime 2026

GenAI has industrialized fraud. Our report maps what investment firms are doing to fight back. With data from 600 senior compliance professionals on AI adoption, EDD automation, and why the firms treating compliance as a growth lever are pulling ahead.

Download the report

EDD as a competitive advantage, not a cost center

The investment firms pulling ahead are the ones that treat EDD excellence as a growth lever rather than a brake. Basic checks, a name run against a static list, or a document accepted at face value, are exactly what GenAI-assisted fraud is built to defeat. Rule-based monitoring and batch-processed screening quickly deteriorate when faced with AI-obfuscated names, addresses, and narratives, generating more false positives even as sophisticated deception slips through. For a sector whose defining risk is PEPs and their networks, that is the difference between a defensible program and an exposed one.

Moving beyond basic checks means richer data and better automation working together. The priority for investment firms is depth: unparalleled PEP and adverse-media data, network and beneficial-ownership analysis, and precision analytics that let experts proactively identify complex risks.

Predictive AI supplements rules-based detection with baseline behavioral modeling, flagging subtle deviations that static systems miss, and is increasingly used to stress-test a firm’s own controls against emerging typologies before criminals find the gaps. Where identity fabrication is the concern, liveness analytics add a further layer by detecting synthetic artifacts in images and audio. Our survey found that among advanced AI users, 54% report increased efficiency, 51% cite an improved customer experience, and 47% note better predictive capabilities.

This combination, efficiency and experience rising together, is the point. Superior data and automation let an investment firm apply genuine scrutiny without imposing friction on legitimate clients. The HNWI who expects a frictionless onboarding gets one; the synthetic identity gets caught. 

For heads of product and growth, that means winning and retaining top-tier clients and growing assets under management (AUM) faster, without compliance acting as the bottleneck. For operations leaders, it means scaling volume without linearly scaling headcount. And for the head of compliance, it means a robust, automated, defensible EDD program built on the industry’s richest PEP data.

This is the exact paradox our report identifies for 2026: AI is simultaneously the primary enabler of financial crime and the most vital tool for controlling it. Investment firms can tilt the scale in their favor by embedding AI quickly and safely, turning high-stakes compliance from a defensive necessity into a durable, competitive advantage.