Anthropic Weighs New AI Model Launch to Counter OpenAI's GPT-6 Astra

Anthropic, the first AI developer to advocate for slowing AI development, is reportedly considering releasing a next-generation model to counter the success of OpenAI's "GPT-6 Astra." The move comes as Anthropic prepares for a year-end IPO. According to Reuters on the 21st, Anthropic has begun safety evaluations for a new model and is debating how to balance investment in development with improving profitability. This follows CEO Dario Amodei's public proposal on the 12th to slow AI development for safety reasons, a stance now challenged by competitive pressure. Anthropic has been a leader in the B2B AI market, but OpenAI is catching up. Data from Ramp shows OpenAI's Astra holds about 13% of corporate AI spending versus Anthropic's Claude at 8%, and OpenAI has overtaken Anthropic in usage costs on OpenRouter for the first time in 2.5 years.