Anthropic warned in its IPO filing that its artificial intelligence models could pose a catastrophic or existential risk to humanity, according to a Reuters report on Monday. The disclosure came as the company prepares a public listing that could raise tens of billions of dollars.

Separately, OpenAI has shelved the release of a new AI model, GPT-6.1 Astra, over safety concerns, BusinessWorld reported. The pause follows internal testing that raised questions about the model’s behaviour.

The two developments landed days after industry CEOs told an Associated Press forum that AI systems need stronger safeguards, with several leaders warning of outcomes they could lose control of. Analysts say the warnings are shifting the debate from capability to containment.

Investors will watch whether the safety disclosures affect Anthropic’s IPO pricing and how regulators respond. For now, both companies say they are committed to third-party evaluations and slower deployment of the most capable systems.