OpenAI, Anthropic Balance IPO Pitch With AI Safety Warnings

According to Axios, OpenAI and Anthropic are trying to reassure Wall Street that their businesses are growing quickly while also telling governments their models do not pose unacceptable risks, as both companies prepare for potentially record-breaking initial public offerings. OpenAI said Tuesday it will soon broadly release its Astra model, but said its most powerful cybersecurity capabilities will initially be limited to trusted testers after the model reached a "critical" cybersecurity threshold. Anthropic, meanwhile, released updated versions of its Fable and Mythos models to address criticism over cost, data sharing and a tendency to refuse legitimate requests. OpenAI warned Astra's safeguards could incorrectly flag legitimate activity as cyber misuse or unauthorized behavior, while Anthropic said its new models are less likely to trigger restricted responses. Anthropic said medical or biology questions will have 85% fewer interventions and some users could see about 60% fewer cybersecurity-related interventions per session. Axios also reported that Anthropic could file a publicly available prospectus as soon as next week, while OpenAI is in earlier stages of its IPO process.