Anthropic is drawing attention on several fronts as it approaches an expected public listing, from the size of an early investor’s stake to cloud capacity pressures and a draft federal AI safety bill.
Former Google CEO Eric Schmidt reportedly owns about 1% of the Claude maker, according to Bloomberg, after investing tens of millions of dollars in its 2021 Series A, when PitchBook data put its valuation at $461 million. That stake is roughly half the 2% reportedly held by co-founder and CEO Dario Amodei. With top-end market estimates valuing Anthropic at up to $2 trillion, Schmidt could add as much as $20 billion to a net worth Bloomberg estimates at $57.5 billion. Schmidt invests through Tomorrow Ventures, Innovation Endeavors Management and his family office, Hillspire, and has also backed AI companies Mistral, General Intuition and Reflection AI.
Meanwhile, a leaked S-1 filing reported by The Wall Street Journal shows Anthropic is struggling to secure enough cloud capacity from AWS, Google and Microsoft, which increasingly view it as a competitor. The filing indicates the company could spend more than $750 billion on cloud contracts, including deals still under negotiation. Anthropic is also leasing its own data centers and working with third-party operators, though it has flagged concerns that their limited operating history could make them less reliable than the hyperscalers.
The filing shows an $80.7 billion net loss in the first half of the year against $16.3 billion in revenue, with two customers accounting for about a fifth of sales. It also warns that liability for damages caused by harmful AI behavior could exceed available insurance. Last week, Anthropic told prospective investors that third-quarter revenue reached $16.8 billion, up 45% from the previous quarter, and the company is reportedly targeting a November IPO.
On policy, Politico reported that Anthropic has been circulating a draft AI safety bill for Congress that would allow a model’s release to be blocked in cases of foreseeable, imminent risk involving more than 50 deaths or over $1 billion in property damage. The risk must arise from a biological weapon, a model evading developer controls, or a cyberattack. These thresholds mirror California’s Senate Bill 53.
Under the proposal, independent evaluators registered with the Commerce Department would assess models and could be paid by AI labs, provided payment does not depend on results. Evaluators identifying an imminent catastrophic risk would have 72 hours to notify the Commerce Department, which would then ask a Washington, D.C., judge to halt the release. No federal agency could stop a model’s distribution on its own. An Anthropic spokesperson confirmed to Politico that the bill reflects the company’s Advanced AI Framework, which it has promoted to federal and state lawmakers since June.
It remains unclear how the system would prevent payments from influencing evaluations. DeepMind chairman Demis Hassabis has previously suggested a self-governing body modeled on the financial industry’s FINRA. Recent industry concern has centered largely on agentic hacking, following incidents in which AI programs disobeyed operators and hacked websites. Anthropic’s proposal nonetheless goes further than others in the industry, including Nvidia CEO Jensen Huang, who has argued that no new AI laws are needed.
