Anthropic Weighs New AI Model Ahead of IPO

The potential launch could come ahead of Anthropic’s anticipated initial public offering. The report said that Anthropic teams are assessing whether to release a system that could respond to OpenAI’s recently deployed GPT-6 Astra. The company is also evaluating safety benchmarks for the proposed model before making a launch decision.

Safety with Competition The possible release comes around two weeks after Anthropic CEO Dario Amodei published a 3,800-word essay calling for a slower pace in the development of advanced AI models . Amodei raised concerns about the safety risks linked to increasingly capable systems. “We must slow the pace at which we improve the capabilities of AI models.

Progress will still seem fast, and we must make wise use of the time we gain,” Amodei wrote. His comments received public support from OpenAI CEO Sam Altman and Elon Musk. Amodei also warned about AI systems gaining greater ability to operate with limited human oversight.

Anthropic is now weighing those safety concerns against pressure from the AI market. The company is reviewing the safety standards of its next system while considering whether to proceed with a launch. GPT-6 Astra Gains Enterprise Share OpenAI launched GPT-6 Astra on September 3 with a focus on autonomous computer use, coding, workplace tools and cybersecurity.

The model reportedly gained traction among corporate users. Data from corporate expense platform Ramp showed GPT-6 Astra accounting for about 13% of monitored enterprise AI spending. Anthropic’s Claude Fable held about 8%, according to the report.

Anthropic’s annualized revenue run rate crossed USD 65 billion by late July, rising from about USD 9 billion at the end of 2025. OpenAI’s annualised revenue run rate passed USD 40 billion in July. Anthropic also faces competition from open-source and open-weight AI models.

Meta is also looking to reduce its reliance on Claude as it expands its internal AI computing efforts.