Anthropic plans new AI model amid intense competition with OpenAI

Reuters reports that Anthropic is considering launching a new AI model during its IPO preparations to counter OpenAI's GPT-6Astra in the enterprise market. Three sources indicate the company is still assessing the safety of the new model and has not finalized a release date. This plan contrasts with Anthropic CEO Dario Amodei's recent call for slowing AI model iterations, as he published a 3,800-word article urging the industry to prioritize safety over rapid development. On September 3, OpenAI released GPT-6Astra, which has drawn attention from enterprise users and prompted some investors to reevaluate Anthropic's market position. Ramp data shows Astra currently captures about 13% of enterprise AI spending, compared to Anthropic's Claude Fable at 8%. As of the end of July, Anthropic's annualized revenue run rate exceeded $65 billion, with projections reaching between $190 billion and $200 billion by 2028. OpenAI's annualized revenue run rate during the same period surpassed $40 billion. Amodei's push for caution highlights the growing tension in the AI sector, as companies balance innovation with safety. The release of GPT-6Astra has intensified competition, with Anthropic seeking to maintain its market share. The potential IPO and new model represent a strategic move to strengthen its position against OpenAI's rising influence in the enterprise AI space.