Mark Zuckerberg told officials at City Hall this week that Meta’s AI agents were not progressing as quickly as leadership had hoped, according to a recording heard by Reuters.
The approval comes nearly two months after Meta cut about 8,000 jobs, or about 10% of its global workforce, and redeployed an additional 7,000 employees to AI-focused teams.
“The trajectory of drug development over at least the past four months has not accelerated as much as we had hoped,” Zuckerberg said, adding that the company’s turnaround bet “has not yet come to fruition.”
He acknowledged that the reorganization was not as “clean” as it should have been and said executives misjudged the timing of the changes. He told staff he expects to see a meaningful return on Meta’s AI investment within three to six months.
Notably, Zuckerberg said he was “very optimistic” about tools like Anthropic’s Claude Code when management began planning the restructuring in January and February earlier this year. These conversations, he said, were driven by fear that the meta wasn’t moving fast enough to keep up. The company is expected to spend up to $145 billion on AI infrastructure this year.
At the same town hall, Meta chief technology officer Andrew Bosworth said an investigation into the company’s suspended mouse-tracking software revealed that no employee data was used for AI training. Bosworth said if the program resumes, it will be done on an opt-in basis, reversing policy from April, when U.S. employees were told there was no way to opt out.
City Hall echoes an earlier internal memo reported by Wired in which Bosworth called the rollout of Meta’s applied AI division “terrible.” Officials of the 6,500-strong force, which was formed in March, described work as chaotic, with one official calling it a “concentration camp.”
Bosworth has pledged to limit the number of managers to 20 direct reports and invest in career development and office perks to rebuild morale.
