Mark Zuckerberg wants you to know that Meta’s future has three pillars. The first two—Family of Apps serving billions of people and an advertising engine reaching hundreds of millions of businesses—are proven revenue machines. The third, announced Monday morning in a four-part X thread and a Meta Newsroom post, is the Meta Enterprise Platform. Zuckerberg called it “the next major pillar of our business.”
As of Monday afternoon, that pillar is a press release.
The initial product lineup—Muse agent, Meta Business Agent, Muse API, and Muse Code—is announced but not launched. Meta has disclosed no pricing, no enterprise launch date, no named customers, and no regional availability. What exists today is a leadership hire, a product list, and a declaration of intent.
What Meta Actually Shipped
The consumer side of the Muse story is real. The personal AI agent launched September 8, 2026, and Sensor Tower estimates (third-party app analytics) put it at 2.8 million downloads in its first two weeks, surpassing ChatGPT as the top free app in the U.S. App Store. Muse handles emails, travel bookings, payments, and shopping. That consumer traction is what makes the enterprise announcement credible enough to pay attention to—but it is not, by itself, an enterprise product.
CNBC reported that Meta hired Chirantan “CJ” Desai as Chief Enterprise Platform Officer, reporting directly to Zuckerberg. Desai was CEO of MongoDB for less than a year. Before that, he was President of Product and Engineering at Cloudflare and, before that, President and COO of ServiceNow. He brings enterprise infrastructure credibility to a company that has historically struggled to hold enterprise customers’ attention.
The market’s read on the announcement was immediate. MongoDB shares fell roughly 18%. Meta’s own stock declined about 4%.
The Trail of Discontinued Products
For enterprise buyers evaluating a new platform, the first question is often not “what can it do?” but “will it still exist in three years?” Meta’s answer to that question is complicated.
The company shut down Workplace, its enterprise communications platform, in June 2026 after a 10-year run. Workplace had roughly 7 million paying users, but growth had been flat since 2021, and it could not compete with Slack or Microsoft Teams. Meta also discontinued Horizon Workrooms, its VR meeting app, in February 2026 as part of Reality Labs cost-cutting. Both products were non-core side projects for Meta, and their closure reinforced a pattern that enterprise buyers notice: when Meta’s advertising business is healthy, enterprise tools get deprioritized.
The trust deficit runs deeper than product longevity. According to an Oppenheimer and Co. survey of 1,500 U.S. consumers that we covered in our Five-Body Problem analysis, only 8 percent of consumers would trust Meta with their passwords. Google scored 30 percent. Apple scored 23 percent. For enterprise IT teams evaluating agent platforms—where the agent will have access to corporate data, internal systems, and business processes—those trust numbers matter.
The Competitive Landscape Meta Is Walking Into
Meta is not entering an empty market. In the same window that Muse was gaining consumer traction, every major competitor shipped enterprise agent infrastructure with identity, audit, and governance baked in.
Microsoft made Entra Agent ID mandatory for new Copilot Studio agents in July 2026, giving every agent its own directory identity. Google’s Gemini Enterprise Agent Platform includes Agent Identity, an Agent Registry, and an Agent Gateway with Model Armor for prompt-injection protection. Salesforce AIforce ships with an AI Control Plane and MuleSoft Agent Fabric for deterministic orchestration. SAP’s AI Agent Hub is vendor-agnostic, inventorying and governing agents regardless of who built them. And NVIDIA just moved agent security into the silicon with its Open Agent Safety Platform and a 120-partner coalition.
As we noted in our analysis of Siri AI’s personal-only launch, the enterprise agent infrastructure race is not about who has the best consumer app. It is about who ships identity management, audit trails, and compliance tooling that enterprise security teams can actually deploy. Microsoft, Google, Salesforce, and SAP have shipped those tools. Meta has announced them.
The Production Gap Behind the Announcements
The timing of Meta’s enterprise push intersects with a broader market reality. A KPMG Q3 2026 survey (fielded July 24-August 25) found that 62 percent of 314 U.S. executives at large enterprises with $1 billion or more in revenue are building, developing, or deploying AI agents. But a vendor-commissioned Cisco survey from March 2026 found that only 5 percent of organizations have reached production, with 60 percent citing security as the primary barrier.
An important caveat: the KPMG figure covers building, developing, or deploying—not just active production. The Cisco survey is vendor-commissioned, which creates an incentive to emphasize security barriers. These numbers are directional, not definitive.
Gartner projects that more than 40 percent of agentic AI projects will be canceled by the end of 2027 due to escalating costs, unclear business value, and inadequate risk controls. That cancellation rate is what happens when enterprises ship agents without the governance infrastructure to manage them—and it is the market Meta is choosing to enter with an announcement rather than a product.
What to Watch For
The deeper forward tell here is not whether Meta can build enterprise tools. They clearly can. The question is whether enterprises will trust Meta enough to deploy them, and whether Meta will commit to this business when the advertising engine remains its primary revenue source.
Watch for three things. First, pricing and launch dates for Muse API and Muse Code. Without those, enterprise developers cannot plan integrations. Second, the first named enterprise customer. Meta Business Agent, announced in June 2026, already claims 1 million pre-launch business adopters across WhatsApp and Messenger—but those are SMB advertisers using Meta’s messaging surfaces, not enterprise IT deployments with governance requirements. Third, the security and compliance architecture. Meta’s announcement says security and privacy are built in from the outset, modeled on the consumer Muse approach. Enterprise buyers will want specifics: audit trails, agent identity management, data residency, and integration with existing identity providers.
Meta has the consumer distribution to make this real. Two hundred million businesses already use its advertising platform. The Muse agent has proven consumer pull. And CJ Desai’s enterprise pedigree—ServiceNow, Cloudflare, MongoDB—signals serious intent. But intent is not a product, and in a market where Microsoft, Google, Salesforce, and SAP have already shipped the governance layer, being announced-but-not-launched is a shrinking window of advantage.