CFTC Commissioner Urges Stronger AI Oversight in Financial Services

CFTC Commissioner Kristin N. Johnson outlined the expanding use of artificial intelligence in financial compliance and enforcement. She warned that without stronger oversight, AI could be exploited to commit fraud and evade regulations.

At a virtual event hosted by George Washington University's Regulatory Studies Center, Commissioner Johnson described how financial institutions are integrating AI tools in compliance, particularly in anti-money laundering ("AML") and sanctions screening. She said that machine learning ("ML") models now analyze millions of transactions to detect anomalies that might elude traditional systems. She noted that these tools are reducing false positives and helping firms identify "red flags early" in both transaction monitoring and internal communications.

Ms. Johnson emphasized the ability of AI to strengthen compliance, but also its potential to facilitate misconduct. She cited findings from a May 2025 GAO report, Artificial Intelligence: Use and Oversight in Financial Services, which identified six core activities where firms deploy AI, including credit decisions, automated trading and fraud detection. She said that firms are "rightly reticent to employ generative AI models" in "use cases that require a high degree of reliability or explainability."

On additional AI risks, she expressed concern over "AI washing"—the marketing of fake or exaggerated AI capabilities to lure investors. She said that "[f]irms may claim to use advanced AI models to generate high returns when, in reality, they rely on rudimentary trading bots or nonexistent systems." She highlighted the CFTC's record-setting $1.7 billion enforcement action against a South African firm that falsely claimed to use a proprietary AI trading bot, calling it a landmark case that illustrates the agency's ability to tackle AI-related fraud with existing tools. (See related coverage.) She warned of challenges around agentic AI, systems that operate with a degree of autonomy. She warned that, "[i]f we integrate agentic AI in compliance, surveillance, and enforcement... we must also examine the limits of synthetic data, ghosts or hallucinations, data leakage, and deepfakes."

Ms. Johnson spotlighted the growing use of supervisory technology—or "suptech"—by federal regulators. She described how the CFTC, SEC, Federal Reserve, FDIC and others are using AI to detect inconsistencies in financial data, uncover outliers and surface early warning signs of misconduct. She also pointed to CFTC's initiatives, including a 2024 Request for Comment and a staff advisory on AI, as steps toward understanding and guiding industry use of these technologies. (See related coverage.)

Ms. Johnson (i) called for heightened penalties for individuals or firms that use AI to intentionally violate CFTC regulations and (ii) reiterated her proposal for an inter-agency task force.