Zuckerberg admits AI agent progress is lagging behind internal expectations at Meta

The race to automate work through AI agents appears to be moving at a slower pace than corporate leaders envisioned, as evidenced by recent comments from Meta's leadership.

During an internal staff meeting this week, CEO Mark Zuckerberg acknowledged that the development of AI agents had not gained momentum at the speed the company had previously forecasted, according to reporting from Reuters.

Meta's AI restructuring gamble#

Earlier in 2026, Meta undertook a major organizational overhaul centered on artificial intelligence. The company terminated approximately 10 percent of its workforce—roughly 8,000 employees—while redeploying approximately 7,000 others into newly formed AI-focused divisions, including a group specifically tasked with what the company calls "Agent Transformation," according to Bloomberg.

Why the restructuring fell short#

Zuckerberg indicated that the executive team had prioritized these drastic moves because they feared the organization couldn't "move fast enough to adapt" to shifting competitive dynamics in technology. However, he conceded that the anticipated benefits of the new AI-centric structure had yet to materialize—at least not at the scale or speed anticipated. The restructuring itself was less efficient than hoped, Zuckerberg noted.

Looking ahead#

Despite the slower-than-expected progress, Zuckerberg expressed confidence that improvements would become visible over the next three to six months as the company's AI investments mature. Meanwhile, the company plans to pour an estimated $145 billion into AI infrastructure this year, Reuters reports.

The reality hasn't matched the optimism: some engineers assigned to Meta's months-old AI unit have described the experience as grueling.