Twenty-six anonymous Meta employees are suing the tech giant, claiming that it used inherently discriminatory AI-powered software systems in a massive round of layoffs. The lawsuit, filed in the Northern District Court of California, alleges that Meta relied on a constellation of internal artificial-intelligence systems to score, rank, and select employees for termination, rather than the considered judgment of managers.
Meta conducted a substantial round of layoffs in May that impacted 8,000 employees, representing 10% of its entire workforce. The layoffs were part of an effort to help offset the hundreds of billions of dollars the tech giant prepares to spend on artificial intelligence development. According to the complaint, the company used an internal large-language model assistant called Metamate, a second brain trained on employee communications and documents, along with algorithmic productivity scores based on keystroke, browser history, and email data. AI-assisted performance review tools and internal records of AI token consumption were also allegedly factored into the decisions.
The plaintiffs argue that these AI systems emphasized metrics like keystroke counts and AI token consumption, which discriminated against employees who had to miss work or produce reduced output due to a disability or protected medical or family leave. When Meta was allegedly made aware of this problem, it did not take the precautions the employees deemed necessary, such as pausing the system for a more neutral review process. The result, the lawsuit claims, was that employees who took protected leaves were disproportionately selected for layoff, based on scoring that not only failed to account for their protected leaves but in effect penalized the employees for exercising their legal rights to these leaves.
Several specific cases highlight the alleged discrimination. One employee was a scientist who was just two days away from giving birth when she was selected. Another, a manager, was on approved pregnancy-related disability leave and became the only person on her team chosen by the system. The plaintiffs include workers on maternity leave, paternity leave, and medical leave for disabilities. One engineer familiar with the February 2025 layoffs said he was aware that employees who took paternity leave had been laid off in that round as well, suggesting a pattern.
Meta denies the allegations. A spokesperson told Gizmodo: These claims lack merit and are not based on facts. Workforce management and organizational decisions were and are made by people, not AI. The company maintains that the layoff decisions involved human judgment, despite the use of AI tools for analysis.
Legal Context and Previous Cases
The lawsuit comes just months after Meta was hit by yet another workplace discrimination lawsuit, this time by a former employee who said older workers were disproportionately targeted in the company’s February 2025 round of layoffs, which impacted 5% of its workforce. At the time, the company said the layoffs were targeting its lowest performers. That case is still pending. The latest complaint builds on concerns about algorithmic bias in employment, a growing area of legal scrutiny. Federal agencies like the Equal Employment Opportunity Commission have issued guidance on the use of AI in hiring and firing, warning that automated systems can perpetuate illegal discrimination if not carefully designed and monitored.
The plaintiffs are now asking the court to block Meta from completing the layoffs on July 22, giving the employees time to pursue claims in private arbitration, as required by their contracts. The workers say Meta’s agreements require employees to arbitrate workplace disputes individually, but do not apply to requests for temporary relief. They are seeking a preliminary ruling from the court blocking Meta from completing the layoffs while they pursue their claims in private arbitration.
Details of the AI Systems at the Center of the Suit
According to the complaint, Meta deployed a sophisticated array of AI tools. Metamate, the internal LLM, was trained on a massive corpus of employee communications, including emails, chat logs, and documents. It could generate performance summaries and rank employees. The algorithmic productivity scores tracked keystroke activity, browser history, and email volume, effectively measuring presence and activity rather than output or quality. AI token consumption—the amount of computing resources an employee used in developing AI models—was also used as a proxy for contribution. Employees on leave would naturally have low or zero scores in these metrics, regardless of their actual value to the company.
The lawsuit alleges that this system created a hidden bias against workers with disabilities, pregnant employees, and those caring for family members. For instance, a worker on short-term disability after surgery would have no keystrokes or AI token usage during recovery, leading to a low score. Similarly, a new parent on parental leave would be absent from the system for weeks or months, effectively ensuring a low ranking. The complaint states that Meta did not adjust the data to account for leave periods or pause the scoring for affected employees.
These AI systems were also used in performance reviews and promotion decisions, further embedding bias into Meta’s talent management. The plaintiffs argue that the combination of Metamate and productivity metrics created a score that was virtually impossible for employees on protected leave to overcome. They claim that Meta knew or should have known about the discrimination because employees raised concerns internally before the layoffs were announced. Despite these warnings, the company proceeded with the AI-driven selection process.
Broader Implications for AI in HR
The case highlights a critical tension: companies are increasingly using AI to cut costs and improve efficiency, but these tools can inadvertently encode biases. AI systems are trained on historical data, which may reflect past discriminatory practices. In Meta’s case, the use of keystroke tracking and token consumption as proxies for productivity is particularly problematic because it penalizes any absence, regardless of reason. Employment law experts note that such systems can violate the Americans with Disabilities Act and the Family and Medical Leave Act, which protect workers from retaliation for taking leave.
Other tech giants have faced similar lawsuits. Amazon was sued for using AI hiring tools that discriminated against women. Uber faced claims over AI-based driver deactivation. The Meta case is notable because it involves layoffs, which are often high-stakes and can affect thousands of workers. If the plaintiffs succeed, it could set a precedent for how courts view AI-driven employment decisions. Companies may be forced to audit their AI systems for bias and provide human oversight to prevent discrimination.
The lawsuit also raises questions about transparency. Employees are often unaware of how AI tools are used in decisions about their careers. In this case, Meta did not disclose the full extent of the AI systems until after the layoffs were announced. The plaintiffs argue that workers should have a right to know the algorithmic basis for their termination. Some states and cities, including New York City and Illinois, have passed laws requiring employers to audit AI hiring tools for bias. California, where Meta is headquartered, is considering similar legislation.
Meta’s massive investment in AI development is central to the context. The company plans to spend hundreds of billions on AI infrastructure, including data centers and chips. The layoffs were framed as a cost-cutting measure to redirect resources toward AI. However, the lawsuit suggests that the very systems Meta is prioritizing may be illegal. The plaintiffs note that using AI to fire people while simultaneously investing in AI is a hypocritical and harmful strategy.
The court has not yet ruled on the request for a temporary restraining order. The hearing is likely to occur before July 22, as plaintiffs urgently seek to stop the layoffs. Meta’s denial of the allegations will be tested in arbitration or court. The outcome will be closely watched by employment lawyers, tech companies, and lawmakers concerned about algorithmic fairness.
One of the engineers in the lawsuit also claims that he was aware that employees who took paternity leave had been laid off in the February 2025 round of layoffs as well. This suggests a pattern that predates the current lawsuit. The engineer says he saw colleagues who were on leave be targeted, and he raised concerns with HR but was ignored. The complaint alleges that Meta’s HR department failed to conduct a meaningful review of the AI system’s impact on protected groups. Instead, the company rushed to implement layoffs to meet financial targets.
The plaintiffs are represented by a firm specializing in employment discrimination and AI accountability. They argue that Meta’s conduct is not only illegal but also undermines trust in the company’s commitment to diversity and inclusion. Meta has publicly stated its support for diversity, yet its AI systems allegedly punished workers for taking family leave. The lawyers plan to introduce evidence of internal communications where executives discussed the potential for bias but proceeded anyway.
Source: Gizmodo News