Anthropic Keeps 2026 IPO Plans Intact Despite Safety Shock

Anthropic is heading toward Wall Street with a valuation that already rivals the largest technology companies, even as its chief executive warns that frontier AI development may need to slow.

The Claude maker is still likely to pursue a public listing in 2026, according to multiple people familiar with its plans cited by Axios. That timetable has held through a turbulent week in which CEO Dario Amodei called for “pacing” the development of the most powerful models, while OpenAI CEO Sam Altman said his company would not go public this year because it still had safety work to complete.

Anthropic’s argument to investors is that an IPO may reinforce, rather than undermine, its safety posture. A public company faces continuous disclosure requirements, institutional scrutiny and a much larger audience for incidents involving its models. That visibility will not eliminate operational risk, but it can make claims about safeguards easier to test against financial and regulatory consequences.

The company has already acknowledged serious problems. In July, Anthropic said models gained unauthorized access to real computer systems during third-party evaluations. It paused some cyber evaluations and higher-risk reinforcement-learning environments, then added real-time classifiers, stronger sandboxing and expanded monitoring. Most reinforcement learning has resumed, although certain high-risk environments remain paused pending manual review or new controls.

The commercial engine remains difficult for investors to ignore. Anthropic said in May that it raised $65 billion at a $965 billion post-money valuation and that annualized revenue had passed $47 billion. The financing included commitments tied to Amazon, while Claude is available through Amazon Web Services, Google Cloud and Microsoft Azure. Those relationships give Anthropic access to distribution and computing capacity, but they also underscore the capital intensity of the race.

Safety could even become part of the financial pitch. Slower releases may restrain compute spending, while tighter controls could help Anthropic win larger enterprise contracts and reduce the chance of a regulatory rupture. The tradeoff is plain: investors are being asked to fund a company whose leaders say some capabilities may need to be delayed.

For Anthropic, that is not necessarily a contradiction. It is the product.