Mark Zuckerberg's reputation for fostering goodwill has taken a hit in recent weeks, with a string of incidents casting a shadow over the Meta CEO. Earlier this month, he faced backlash after his superyacht left a smaller boat’s passengers stranded in the frigid waters of Alaska. Adding to the turmoil, Meta agreed to a staggering $17 billion settlement in response to allegations from several states that its platforms, Instagram and Facebook, pose risks to children and teenagers.
In a further twist, revelations surfaced about the company's plan to initiate mass layoffs, intending to replace a significant number of its workforce with AI systems, leading to fears of a dystopian workplace environment. This plan reportedly faced internal pushback, as technology struggled to meet high expectations.
According to reports from Reuters, during a leadership retreat held in January at Zuckerberg’s lavish $100 million retreat in Hawaii—often dubbed a “postapocalyptic bunker”—the Meta founder and executives devised an ambitious strategy termed Project OT. The goal was to transform Meta into an “AI native” organization, wherein AI would assume numerous roles traditionally filled by human employees. Smaller teams of humans were envisioned to manage the AI systems that would take over many daily tasks. However, this restructuring meant substantial layoffs, with discussion of potentially reducing headcount by up to 60% in some departments.
By May, after a prior wave of layoffs eliminated 10% of the workforce, it was reported that employee numbers in certain engineering teams had already decreased by 30%. Compounding frustrations arose when it was revealed that Meta had implemented tracking software on employees' devices to monitor their keystrokes and clicks, intended to train AI for human-like interactions. This led to anxiety among staff who felt they were unwittingly mentoring their potential replacements, driving many to vent their frustrations on the company’s internal communication platform.
Employee morale plummeted following these developments, with favorable sentiment dropping from 74% to 55%, prompting an uptick in labor organizing activities. Internal investigations revealed that the AI systems central to Project OT were underperforming. Although there was an uptick in code creation from staff utilizing AI, actual productivity remained ambiguous. Teams expressed concerns early on about the “reliability warning signs” emerging from the surge in AI-generated code, indicating significant issues.
The public caught a glimpse of the bigger picture when hackers exploited a vulnerability in Meta's new AI-enhanced customer support bot, leading to unauthorized access to prominent Instagram accounts, including the once-active Obama White House page.
Following the initial round of layoffs in May, Zuckerberg assured remaining employees that no further company-wide layoffs were anticipated for the remainder of the year. In subsequent weeks, he made pledges to enhance workplace amenities and openly acknowledged errors during a town hall meeting. A video campaign emphasized Meta's commitment to empowering its workforce. Additionally, Zuckerberg shared a lengthy manifesto outlining the company’s dedication to using AI technology to benefit users rather than solely automate jobs.
In a statement to Reuters, Meta confirmed the earlier plans for layoff waves and the potential for team reductions of up to 60%. The company explained that their scenario planning aimed to explore various outcomes, including redeployments and team closures. While many employees were reassigned to critical new roles, the company ultimately chose not to implement every scenario explored.
However, skepticism remains among employees, with Zuckerberg’s language regarding “company-wide” layoffs hinting to some that further cuts might still be on the horizon, either in a targeted manner within specific teams or across the company in the next year.
