Zuckerberg tells staff Meta’s AI agent progress is “slower than expected,” after 8,000 layoffs

Meta CEO Mark Zuckerberg acknowledged internally that the pace of Meta’s AI agent work has not increased as much as leadership expected over the prior four months, according to a recording of a town hall session heard by reporters and later summarized by news outlets. The comments came shortly after the company carried out a broader restructuring that included layoffs affecting about 8,000 employees, Reuters said in a separate report on July 2.

In the account, Zuckerberg told employees that AI agent development had “not really accelerated in the way that we expected,” and described the results as falling short of what the internal team expected to see after aggressive changes. The remarks were reportedly made during an internal meeting dated July 2, 2026, according to outlets describing the same town hall discussion.

The reports characterize Meta’s admission as an internal adjustment to expectations around a product and platform shift toward so-called AI agents, software systems designed to take actions toward a goal rather than only responding with text. Meta has described agents as a major part of its artificial intelligence push, but the town hall comments suggested that progress against internal milestones has been slower than planned.

Multiple summaries said Zuckerberg’s comments also reflected a broader reevaluation of the effectiveness of the company’s restructuring, which included reducing headcount and reorganizing teams in an effort to refocus resources. While the details of what specifically changed, which teams were responsible, and what engineering or product bottlenecks remain were not disclosed in the public summaries.

One frequently cited figure in the coverage is a reported “$145B bet” tied to Meta’s 2026 AI spending and agent development efforts, which Tech Times said Zuckerberg referenced in his discussion with staff. The figure has not been confirmed in the material available for this write-up, so it should be treated as reported rather than independently verified.

Company context underscores why the admission drew attention. For Meta, AI agents are positioned as both a new computing layer for consumer experiences and a way to support advertising, commerce, and other business workflows. If agent capability is arriving later than expected, it can affect product timelines, developer tooling rollout, and investor confidence in the durability of the company’s AI strategy.

Meta did not provide a public statement responding to the internal remarks in the information available here. As a result, the reports leave open key questions, including what measurable targets were missed, whether the pace improved in certain subprojects, and whether Meta plans any immediate changes to staffing, budgets, or release schedules tied specifically to agent development.

Going forward, investors and employees will likely watch for indicates of execution in publicly visible milestones, such as new agent-enabled features across Meta’s apps and updates to AI infrastructure. Another item to track will be whether Meta’s reported internal messaging leads to changes in how the company describes timelines for agent rollouts in future earnings commentary or product updates.