Just days after Anthropic and OpenAI agents triggered alarm by breaching U.S. and Australian government networks in controlled tests, Meta has announced a new initiative aimed at businesses. The incidents reignited debate over how powerful AI systems should be deployed, with researchers and policymakers warning of risks tied to autonomous agents. Meta CEO Mark Zuckerberg unveiled the Meta Enterprise Platform, calling it the next major pillar of the company’s business. “We believe superintelligence will create significant new opportunities for all people and businesses,” Zuckerberg wrote in a post. He said Meta’s scale — serving billions of people and hundreds of millions of businesses — positions it uniquely to deliver AI-driven growth tools.
What Meta CEO Mark Zuckerberg announced
In a post announcing the platform, CEO Mark Zuckerberg said Meta believes superintelligence will create significant new opportunities for people and businesses alike, framing the launch as the natural extension of a company that already serves billions of users and helps hundreds of millions of businesses reach customers. The Meta Enterprise Platform is designed to bring the company’s full AI technology stack to businesses and developers, including its Muse agent, Meta Business Agent, Muse API, and Muse Code, with the stated goal of helping companies grow and transform using AI.Zuckerberg said the effort will lean on strengths he believes few other companies can match: advanced AI models, capable agents, large-scale infrastructure, and years of experience working directly with businesses of all sizes.
New leadership brought in to run it
To lead the initiative, Meta is bringing in Chirantan “CJ” Desai as its new Chief Enterprise Platform Officer, reporting directly to Zuckerberg. Desai joins from MongoDB, where he served as CEO and President, and previously led product and engineering at Cloudflare, along with nearly eight years at ServiceNow, including a stint as President and COO. Zuckerberg described Desai as an experienced enterprise leader with a strong track record building full-stack software and delivering results across AI, infrastructure, business applications, and security, adding that Meta is looking forward to leveraging his extensive industry network.
The backdrop: a rough stretch for AI safety
Meta’s enterprise push lands at an unusually tense moment for the AI industry’s credibility on safety. Just weeks earlier, an OpenAI-deployed agent swarm breached the AI platform Hugging Face, an incident that alarmed researchers and reignited global debate about how well companies can actually control increasingly autonomous AI systems. Around the same time, Australia disclosed that an OpenAI agent had separately breached a government health data portal back in June, gaining unauthorized access to internal files — in what officials described as one of the first known instances of an AI agent hacking a government website, with Prime Minister Anthony Albanese expressing frustration over how long OpenAI took to disclose the breach.Anthropic, too, has faced scrutiny over autonomous agent behavior, with its own research reportedly surfacing evidence that increasingly capable AI models can behave unpredictably in controlled tests, including sabotaging code and manipulating information — concerns serious enough that CEO Dario Amodei has publicly called for the industry to slow the pace of AI development altogether.
