The United States Securities and Exchange Commission has put AI companies on notice in plain terms. On 30 October 2026, SEC Enforcement Director David Woodcock told Reuters that the agency is examining artificial intelligence companies' corporate disclosures as investor worries mount over a potential AI bubble that may have been fuelled by circular deals. The agency does not want to stand in the way of innovation, Woodcock said, but where disclosure falls short, enforcement will be present. For founders who describe their products as AI-powered, that sentence is the story.
What the SEC actually said
Woodcock, who became SEC enforcement director in May 2026, said the agency's principal concern is that people are making accurate disclosures about material information so that investors know what they are getting into. "Our principal concern is that people are making accurate disclosures about material information so that investors know what they're getting into," he said in the interview. Where such disclosure does not happen, he added, "then enforcement will be present."
What "AI washing" means in securities law
AI washing occurs when a business exaggerates or misrepresents its artificial intelligence capabilities, adoption, or financial benefits. Using AI-related language is not inherently misleading. The problem arises when companies make materially false statements, or omit information investors need to evaluate those claims. Materiality is the key test: would a reasonable investor consider the information important in deciding whether to buy, sell, or hold a security? A vague claim that a company "uses AI" may be puffery. A specific claim that a proprietary model drives revenue, when the revenue actually comes from manual work or a third-party tool, is a different matter.
The SEC has already pursued such cases. In March 2024, regulators charged investment advisers Delphia and Global Predictions with making misleading statements about their use of AI. The firms agreed to pay a combined $400,000 in civil penalties without admitting or denying the findings. Delphia claimed to use customer data and machine learning to make investment decisions, despite lacking the advertised capabilities. Global Predictions similarly exaggerated its AI-powered forecasting services. Another case involved Rimar Capital, whose representatives raised nearly $4 million from investors while misrepresenting an AI trading platform, with $310,000 in combined civil penalties announced in October 2024. Those cases show that securities laws apply to AI claims just as they do to claims about revenue, products, or investment performance.
Circular deals and private markets are in scope too
Woodcock said the SEC will continue to look at issues in private markets, adding that the enforcement team's asset management unit is very focused on conflicts of interest and valuations. "When people don't make accurate disclosures, whether it's about valuations or whatever, then there is a role for enforcement," he said. Reuters previously reported the SEC was probing a type of private equity fund as part of the agency's broader interest in opaque markets. Woodcock also said enforcement staff are working with policy officials as well as the Commodity Futures Trading Commission, which broadly oversees swap markets, to monitor and regulate the booming new markets.
A practical disclosure audit for founders
The practical response, as law-firm and practitioner commentary has framed it, is an evidence file: for every AI claim a company makes, a record of what the system does, how it was tested, and what could go wrong. Start with an inventory of every public AI claim, in investor decks, filings, press releases, product pages, and earnings calls. For each claim, ask four questions. What does the system actually do today, as opposed to what is on the roadmap? What data, models, and human processes are involved? What are the material risks, costs, and dependencies, including compute spending, supplier concentration, and model limitations? And who approved the claim, on what evidence?
Why it matters for founders and investors
For founders, the lesson is to treat AI language as a liability surface, not just a marketing asset. Every "AI-powered," "proprietary model," and "autonomous" claim should survive a simple test: can you show an investor, a regulator, or a journalist the system, the test results, and the limitations behind the words? If the answer is no, soften the claim before you publish it, not after the SEC asks about it. Early-stage companies are not exempt from this discipline in spirit, even if they are not yet public filers: private investors, acquirers, and future underwriters will all read today's deck against tomorrow's diligence.
For investors, the interview is a diligence prompt. Ask for the evidence file behind the AI story. Ask how much revenue depends on AI features versus services, how much spending is committed versus discretionary, and whether any major customer is also an investor, supplier, or strategic partner. Ask about valuation methods in private rounds, and about conflicts of interest in funds that hold both sides of a circular deal. The companies that answer those questions cleanly are not merely safer; they are more fundable, because they reduce the uncertainty that is currently driving the AI valuation debate.
Angeled.com connects founders with investors.
Read Next
Open Weights Are Not Open Business Models: Mistral Large 4 and the Licensing Trap
Courts Are Writing AI's Rules Now: Three Rulings Founders Should Read
Delhi High Court Orders 24-Hour Takedown of AI Deepfake Videos Targeting Chief Minister
EU Court Adviser Says AI Training Is Legal — but Gemini's Answers May Still Infringe Copyright
AI Regulation After the Midterms: What Founders Should Do Before the January Bills Arrive
DOJ Launches Super Intelligence Task Force to Police AI Misuse



