Meta has appointed Chirantan "CJ" Desai, previously the chief executive officer of MongoDB, to the new role of Chief Enterprise Platform Officer, reporting directly to CEO Mark Zuckerberg. Desai will lead the "Meta Enterprise Platform," a division focused on offering AI services tailored for businesses. This initiative aims to help companies use artificial intelligence to drive growth and innovation, aligning with Zuckerberg's vision of making advanced AI accessible to individuals, businesses, and developers. The platform is expected to compete with similar offerings from major tech companies like Microsoft, Google, Amazon Web Services (AWS), and AI-focused firms such as OpenAI and Anthropic. Among the early products under the Meta Enterprise Platform are the Muse Agent, a professional version of a consumer AI tool; the Meta Business Agent for merchants; the Muse API; and the Muse Code coding tool. However, Meta has not yet provided details on the specific features, pricing models, or security guarantees of these products. MongoDB confirmed that Desai will leave the company immediately. The board of directors has brought back Dev Ittycheria as interim CEO, while also hiring a recruitment firm to find a permanent replacement. Ittycheria previously led MongoDB from 2014 until last year, during which the company's annual revenue grew from about $35 million to over $2.3 billion. MongoDB's stock price dropped sharply after the announcement of Desai's departure to Meta. Meta introduced an AI-powered personal assistant called Muse in early September 2026, designed for users aged 18 and older who need help with daily tasks. Muse runs on a secure virtual machine and can be accessed through a dedicated app or via WhatsApp. A business version of the AI, the Business Agent, has already been used by more than one million merchants on WhatsApp and Messenger. In August 2026, Meta revealed that the Muse Spark 1.1 AI model had accessed the internet and breached the systems of a third-party company during cybersecurity tests. This breach was caused by a configuration error from the cybersecurity provider Irregular. With this disclosure, Meta joined OpenAI and Anthropic as the third major AI company to admit that one of its models had compromised external systems during testing.