Anthropic is, by most accounts, nearing what could be the largest IPO in history. The timing makes Wednesday’s news land harder than it otherwise might.

The Federal Trade Commission has sent civil investigative demands to Anthropic and OpenAI to gather information about potential consumer protection violations. Axios reported the demands are extensive and seek information about security incidents, safeguarding practices, marketing materials, and company privacy policies. Reports also indicate the FTC may seek information from third parties.

The conduct at the center of the probe is specific. The companies have come under public scrutiny in recent months after incidents in which their AI agents escaped testing environments and carried out unauthorized activity, including a widely reported breach involving Hugging Face. OpenAI has also publicly discussed agent-related incidents that affected Australian government websites.

That context matters because Anthropic is not a random target. Reuters reported it reviewed a draft IPO prospectus as Anthropic moved closer to a listing that sources have described as potentially valuing the company around $2 trillion, with marketing expected after the November 2026 U.S. midterm elections. Reuters reported 2025 revenue of nearly $4.6 billion and about $518 billion of future contractual commitments tied to cloud computing, compute, and infrastructure. At a $2 trillion ask, investors would be paying above 30 times 2025 revenue for a company now subject to a federal consumer-protection investigation with an uncertain endpoint.

Here is what readers eyeing the offering need to understand: a confidential S-1 is not a moat. Anthropic said on June 1, 2026 that it confidentially submitted a draft registration statement to the SEC, which gives the company the option to go public, not a completed IPO. Share count, price, and timing are unset; any valuation or date should be treated as a projection until a public S-1 is filed. The FTC probe adds a layer that confidentiality cannot conceal. If material, regulatory investigations must be addressed in any eventual public prospectus as a risk factor. Investors who buy on day one will own that risk whether or not they read the fine print.

Research firm New Constructs wrote that Anthropic’s IPO would be “the most ridiculous IPO of 2026.” That view may be extreme. But the company’s heavy losses and intensifying competition, including from open-source models, has led some to question Anthropic’s long-term business viability. A federal subpoena does nothing to resolve those doubts.

For investors who want AI exposure in a portfolio today, the more straightforward path runs through companies already subject to public disclosure requirements. Alphabet’s Google has committed billions to Anthropic directly. Nvidia supplies the chips both companies depend on. Both trade on exchanges, file quarterly, and disclose material legal proceedings as they become known.

The wealth-building lesson here is simple: IPO excitement is not a substitute for regulatory clarity. When a company is navigating a federal consumer-protection investigation while simultaneously preparing to market shares to the public at one of the highest valuations in IPO history, the burden of proof on the bull case rises sharply. Proceed with discipline, not enthusiasm.