Two acquisitions on the same day reveal the different anxieties of the AI giants before their IPO: OpenAI buys a microphone, Anthropic buys a laboratory.

According to 1M AI News monitoring, OpenAI and Anthropic each completed an acquisition on the same day, but the direction of the deals could not be more different. OpenAI bought the tech podcast TBPN, while Anthropic acquired AI biotech company Coefficient Bio for about $400 million. The logic differences between the two transactions happen to reveal each company’s most urgent shortcoming as they count down to an IPO.

《The Information》 reported the complete backstory behind the TBPN deal. The acquisition was personally driven by Fidji Simo, CEO of OpenAI’s AGI deployment division. It began with the company’s streak of public relations missteps over the past few months: the lawsuit response to allegations that ChatGPT encouraged users to self-harm was deemed inappropriate, CEO Sam Altman’s comment on a late-night talk show—“You know, you can’t have ChatGPT but you don’t know how to raise a child”—was seen internally and publicly as out of touch, and former Chief Communications Officer Hannah Wong has also left. Simo started looking for a new approach to public communications earlier this year. After the news was released, some OpenAI employees thought it was a long-overdue April Fools’ joke.

Simo joined OpenAI in August last year after serving as CEO of Instacart for four years. Over the past seven months, she led a series of decisions to “cut off side missions”: shutting down the video generation tool Sora, scaling back ChatGPT’s instant checkout feature, shelving plans for sexual content, and canceling quarterly planning in favor of doing it only once every half-year. Because she has long suffered from postural orthostatic tachycardia syndrome, she has largely been working remotely in Southern California. In recent weeks, her condition has flared up repeatedly, causing her to miss some meetings. But her appearance frequency in investor presentations has clearly increased, and several people with knowledge of the matter believe the company’s power center is shifting from Altman to Simo. At the same time, at least three teams inside OpenAI are developing products that are similar competitors to Anthropic Claude Cowork.

On the other hand, Anthropic chose a completely different path. Over the past six months, it filled key pieces of the Claude ecosystem through three acquisitions: in December 2025, it acquired the JavaScript runtime Bun to strengthen the Claude Code foundation; in February 2026, it acquired computer operating company Vercept to add computer use capabilities; and in April, it acquired Coefficient Bio to move into life sciences. In parallel, its product lines are clearer: in October 2025, it launched Claude for Life Sciences, connecting research tools such as Benchling, BioRender, and PubMed, with built-in Agent Skills that can automatically draft experimental protocols and process genomic data; in January 2026, it expanded to Claude for Healthcare, opening it to doctors and insurance organizations in a HIPAA-compliant environment. Partners include multinational pharma companies such as Sanofi, Novo Nordisk, and AbbVie. Coefficient Bio fills the layer Anthropic has not yet built on its own: enabling AI not just to connect to existing research tools, but to directly plan the entire drug R&D workflow.

By comparison, the anxiety of the two companies points in different directions. About 60% of OpenAI’s revenue comes from the consumer side, and public goodwill directly affects its valuation narrative. Buying TBPN, at its core, is buying a narrative machine that can run continuously. Anthropic’s focus is the enterprise market, but enterprise customers need to see that the vendor has an irreplaceable depth of integration within its own industry. By acquiring Coefficient Bio and building the Claude for Life Sciences product line over six consecutive months, its goal is to establish switching costs within pharma companies’ R&D pipelines. An Axios analysis article published the same day likened the competition to a “capital allocation game,” and both companies plan to IPO later this year, meaning their weaknesses must be addressed before roadshows.