California Attorney General Rob Bonta has served an investigative subpoena on OpenAI. The legal action targets the company's handling of recent cybersecurity incidents.

This development marks a significant escalation in state-level regulatory oversight of artificial intelligence firms. The subpoena requires OpenAI to provide documents and information related to specific security breaches or vulnerabilities identified by the Attorney General's office.

Regulatory context

The investigation is part of an ongoing review by the California Department of Justice. State authorities are examining whether the AI developer complied with consumer protection laws and data security standards during the incidents in question.

No financial penalties or specific incident dates were disclosed in the initial press release from the Attorney General's office. The focus remains on gathering evidence to determine if statutory violations occurred.

Investor Michael Burry stated that markets should "tank hard" to prevent initial public offerings by OpenAI and Anthropic. He argued this action is necessary "for the benefit of humanity" to stop these companies from absorbing and destroying trillions in capital.

Burry's Warning on AI Capital

Burry took to social platform X to voice his concerns about the artificial intelligence sector. He responded to a user suggesting that both companies could lose value post-listing by asserting they would "suck up and then destroy TRILLIONS of dollars of capital." He added that such financial damage would be the least of the harm caused.

The investor has previously compared current enthusiasm for AI stocks to the late-1990s dot-com bubble. He has warned that investors may be overstating the earnings potential of technology companies.

IPO Plans and Financial Disclosures

OpenAI confidentially submitted a draft registration statement for a potential IPO in June but has not set a listing date. CEO Sam Altman indicated the company is not planning to go public in 2026. Reports suggest OpenAI is seeking a $30 billion funding round at a valuation of about $1.4 trillion.

Anthropic also confidentially filed for an IPO in June. Recent reports point to a potential November listing with a valuation around $2 trillion, though terms remain unfinalized.

What the Numbers Show

Anthropic's disclosures reveal a stark divergence between revenue generation and capital expenditure. The company reported $4.6 billion in 2025 revenue against operating losses exceeding $8 billion. This indicates that current operational losses are nearly double the annual revenue. Furthermore, the company outlined hundreds of billions of dollars in future computing commitments, suggesting that the cash burn rate will remain high relative to income as it scales infrastructure.