Zuckerberg admits ‘mistakes’ in Meta’s AI overhaul

Zuckerberg Admits Mistakes in Meta AI Transformation Memo

Mark Zuckerberg acknowledged on Thursday that Meta has made mistakes during its push to rebuild the company around artificial intelligence, according to an internal memo seen by Reuters.wired

“Given the complexity of these changes, we’ve made mistakes and will almost certainly make more,” Zuckerberg wrote, adding that he is focused on providing “as much stability as possible” going forward.usnews

The memo also indicated that Meta plans to explore new roles for employees who have been reassigned to help train AI models, and that Zuckerberg does not anticipate further company-wide layoffs for the rest of the year.streetinsider

A Turbulent Transformation

The admission comes weeks after Meta executed one of its largest restructurings to date. On May 20, the company laid off approximately 8,000 employees — roughly 10 percent of its workforce — while simultaneously reassigning 7,000 others into AI-focused roles. The combined changes affected about 20 percent of Meta’s staff.cnbc

In his earlier memo on May 20, Zuckerberg warned employees that “success isn’t a given” in the AI race and acknowledged that the company had not been as transparent as it should have been. The layoffs were driven by the mounting costs of Meta’s AI infrastructure spending, which the company has projected at up to $145 billion through 2028.reuters

Low Morale and Internal Criticism

The turmoil inside Meta had been building for weeks before Zuckerberg’s latest acknowledgment. A Wired report published in May, based on conversations with more than a dozen current and former employees, described the atmosphere at the company as “alarmingly and historically dismal” ahead of the layoffs.wired

Employees had signed petitions against being monitored by AI tools, and the reshuffling of thousands into unfamiliar roles left many uncertain about the company’s direction. Business Insider reported that Zuckerberg’s softer tone in recent communications reflected a deliberate strategy to reassure a rattled workforce, with the CEO emphasizing his desire to provide stability to those who remained.nytimes

Broader Context

Meta’s AI strategy has faced scrutiny for months. In March, the New York Times reported that the company had delayed its next-generation AI model, codenamed Avocado, after it failed to match the performance of rivals from Google, OpenAI, and Anthropic in internal benchmarks. Reuters had separately reported in March that Meta was contemplating cuts affecting up to 20 percent of its workforce as AI costs mounted.nytimes

Zuckerberg’s latest memo, with its frank admission of error, signals an effort to reset expectations internally even as the company continues to pour resources into a bet whose payoff remains uncertain.