Meta CEO Mark Zuckerberg has admitted that the company’s ambitious push into AI agents has not progressed as quickly as executives initially anticipated, raising fresh questions about the effectiveness of the social media giant’s recent restructuring efforts and massive artificial intelligence investments.

Speaking during an internal company town hall meeting, Zuckerberg acknowledged that Meta’s expectations for AI-powered agents over the past several months have not materialized as planned, despite significant organizational changes and billions of dollars invested in artificial intelligence infrastructure.

Zuckerberg Says Meta’s AI Agent Strategy Has Yet to Deliver Expected Results

According to comments made during the meeting, Zuckerberg revealed that the company’s leadership team expected AI agent technology to accelerate much faster during the first half of 2026.

AI agents are advanced software systems designed to perform tasks autonomously on behalf of users, and they have become one of the most competitive areas in the artificial intelligence industry.

Reflecting on Meta’s recent progress, Zuckerberg admitted that the company’s restructuring strategy and AI investments have not yet produced the expected breakthroughs.

“The trajectory of agentic development over the last few months hasn’t accelerated in the way we expected,” Zuckerberg reportedly told employees.

He added that Meta executives were highly optimistic earlier this year, particularly after observing rapid advancements from AI companies developing agent-based systems.

Massive Workforce Changes Were Intended to Accelerate AI Development

Meta implemented one of its largest internal reorganizations earlier this year as part of its broader artificial intelligence strategy.

The company reportedly reduced its workforce by approximately 10% globally while moving around 7,000 employees into AI-focused roles. The restructuring was designed to free up resources for expensive AI infrastructure projects and position Meta as a leader in next-generation artificial intelligence.

However, Zuckerberg admitted that the transition process was not handled as smoothly as originally planned.

Meta’s Recent AI Restructuring Efforts

Area	Details
Workforce Reduction	Approximately 10% of global staff
Employees Reassigned	Around 7,000 workers moved to AI teams
Primary Goal	Accelerate AI innovation and efficiency
Current Status	Results slower than expected
Future Outlook	Improvements expected within 3 to 6 months

The changes reportedly sparked internal concerns among employees regarding job security, morale, and the pace of organizational transformation.

Meta Still Plans to Invest Billions Into Artificial Intelligence

Despite acknowledging slower-than-expected progress, Zuckerberg remains optimistic about Meta’s long-term AI strategy.

He told employees that the company expects to begin seeing more meaningful returns from its AI investments within the next three to six months.

Meta is projected to spend up to $145 billion on artificial intelligence infrastructure during 2026, making it one of the largest AI investment programs in the technology industry.

The spending forms part of a broader race among major technology companies to dominate artificial intelligence development, cloud infrastructure, and autonomous AI systems.

---

Meta May Relaunch Employee Mouse-Tracking Program With Opt-In Controls

During the same internal meeting, Meta Chief Technology Officer Andrew Bosworth addressed concerns surrounding the company’s controversial employee activity monitoring software.

The software, which tracked employee mouse movements and other digital interactions, was introduced earlier this year to help train artificial intelligence systems and improve productivity analysis.

However, the initiative faced criticism after reports surfaced that sensitive information may have been exposed during the data collection process.

Internal Investigation Finds No Employee Data Was Used for AI Training

Bosworth confirmed that Meta’s ongoing review found no evidence that employee data collected through the monitoring program had been incorporated into AI training models.

The company suspended the program last month while conducting its internal investigation.

Timeline of Meta’s Employee Monitoring Controversy

Event	Status
Employee monitoring software introduced	April 2026
Employee concerns emerge	Shortly after launch
Program suspended	June 2026
Internal review conducted	Ongoing
AI training misuse found	None reported
Potential relaunch	Possible on an opt-in basis

Future Participation Could Become Voluntary

Bosworth revealed that if the monitoring system returns, employees will likely have the option to choose whether they want to participate.

He explained that workers comfortable with contributing data could continue doing so, while others would be free to decline without consequences.

The move represents a significant change from Meta’s original rollout of the software, when employees reportedly had no option to opt out.

The developments highlight the growing challenges facing major technology companies as they balance aggressive artificial intelligence development with employee trust, workplace privacy concerns, and investor expectations.