Zuckerberg frames Meta’s enterprise AI push as bigger than just “agents”

Meta’s AI pitch to businesses goes beyond building “agents,” CEO Mark Zuckerberg said during the company’s second-quarter earnings call on Wednesday, arguing that the enterprise opportunity includes the software and infrastructure businesses need to connect AI to workflows.

According to coverage of the call, Zuckerberg described a “large enterprise opportunity” that stretches across multiple layers of a future AI stack: AI agents, application programming interfaces (APIs, which are standardized software interfaces that let one system call another), compute (the computing resources required to run AI), and Meta’s own internal software.

The framing is notable because many enterprise AI conversations currently focus on agent-like tools that can take actions on a user’s behalf, such as drafting, searching, or completing multi-step tasks. Zuckerberg’s comments suggested that Meta wants to sell something more durable to enterprises: connectivity and operating infrastructure around those tools, not only the agent experience at the top.

Meta did not, in the coverage summarized here, disclose specific contract announcements, customer names, or quantified revenue outlook tied to this enterprise strategy. Nor did it provide segment-level breakouts that would show how much of the company’s earnings trajectory is linked to enterprise AI adoption.

The discussion also underscores how Meta is trying to position its AI efforts as both a product and a platform play. By emphasizing APIs and internal software, Zuckerberg implied that Meta believes enterprise buyers will care about integration effort, reliability, and performance, in addition to the capabilities of any single chatbot or agent.

In the broader technology sector, enterprise AI is increasingly moving from isolated pilots toward systems that can be embedded in business processes, with emphasis on security, governance, and the ability to connect models to existing applications. Meta’s approach, at least as characterized in this earnings-call coverage, aligns with that shift by describing enterprise value as a combination of model-driven automation and the integration layer around it.

Still, there is limited detail in what has been shared from the call. The reported comments do not specify which Meta AI agent offerings or APIs are being prioritized, what enterprise industries are the target first, or what timeline the company is working toward. For now, the clearest takeaway is the scope of Meta’s ambition rather than the pace of monetization.

What to watch next is whether Meta follows up with additional disclosures in future earnings materials, including any clearer metrics on enterprise usage, pricing, partnerships, or early customer deployments. Investors and business buyers will likely look for signs that the company can convert an enterprise platform narrative into measurable demand, rather than keeping it as a high-level opportunity statement.