Twenty-six current and former Meta employees have filed a federal lawsuit accusing the company of using artificial intelligence systems that disproportionately targeted workers with disabilities or who had taken medical leave when selecting people for mass layoffs. The suit, filed Monday in Oakland, California, alleges that Meta relied on productivity metrics and AI token usage data when it cut roughly 8,000 jobs beginning on May 20, effectively penalising employees who had missed work for protected reasons. It appears to be the first lawsuit against a major US technology company to challenge the use of AI in conducting layoffs.
The Allegations in Detail
The 71-page complaint names a constellation of internal AI systems that the plaintiffs say Meta used to score, rank, and select employees for the termination list. Those included Metamate, a large language model assistant, along with employee-trained “second brain” agents, keystroke and activity monitoring data, AI token usage dashboards, and algorithmically assisted performance ranking. The plaintiffs argue that because the system relied on output metrics, it structurally disadvantaged anyone who had taken time away from work and therefore had fewer data points to measure.
All 26 plaintiffs filed anonymously and say they share one thing in common: each took, requested, or was approved for protected leave within the past 24 months. They come from six states including California, New York, Florida, Illinois, Pennsylvania, and Washington, plus the District of Columbia. Their claims include violations of the Americans with Disabilities Act, the Family and Medical Leave Act, the Pregnancy Discrimination Act, and state-level leave protections.
The lawsuit also alleges that Meta deployed a monitoring program earlier in 2026 that captured keystrokes, screen content, mouse activity, browser history, messages, emails, and location data from company-issued devices. The program was announced through what the complaint describes as a low-visibility internal post by an engineer rather than a senior leader, with no consent prompt on at least some teams. Data from the monitoring program was used to build AI tools, including by a new engineering organisation to which employees were reassigned on a mandatory basis, according to the complaint.
Legal Context and Precedent
This case arrives as courts and regulators around the world grapple with how existing employment law applies when companies use algorithmic systems to make or inform decisions about who stays and who goes. In the United States, the Equal Employment Opportunity Commission (EEOC) has issued guidance warning that AI-powered hiring and firing tools must not discriminate against protected groups under Title VII of the Civil Rights Act, the ADA, and the ADEA. However, enforcement has been limited, and private lawsuits like this one are beginning to fill the gap.
The plaintiffs are seeking a preliminary injunction to block their terminations, which are set to take effect on July 22, while they pursue individual claims in arbitration. They also allege that Meta failed to test its AI systems for bias in violation of recently adopted California and New York City laws that regulate AI in employment decisions. California’s AI accountability law, which took effect in 2023, requires companies to conduct bias audits on automated decision-making tools. New York City’s Local Law 144, enacted in 2022, mandates independent audits of AI hiring tools. Meta’s alleged use of AI in layoffs could fall under these statutes, though the legal landscape remains unsettled.
Judge William Orrick in the Northern District of California has been assigned to the case. Orrick has presided over several high-profile tech employment disputes, including class actions against Uber and Lyft regarding worker classification. His rulings on motions to dismiss and class certification will be closely watched by both tech companies and employment attorneys.
Meta’s Defense and Broader Implications
A Meta spokesperson said the claims lack merit, stating that workforce management and organisational decisions were and are made by people, not AI. “We take our legal obligations seriously and have always made human-led decisions about performance and employment,” the spokesperson said. “These allegations are based on a misunderstanding of how our tools work.” However, the lawsuit provides detailed internal documents and email trails that suggest AI outputs influenced termination lists, with managers reportedly instructed to use the scoring dashboards as a “starting point” for decisions.
The case raises profound questions about the role of automation in the workplace. As companies increasingly deploy AI systems to track productivity, monitor communications, and assess employee performance, the risk of algorithmic discrimination grows. Workers with disabilities, pregnant employees, those on family leave, or individuals with chronic illnesses may be disproportionately impacted because their activity metrics dip during absences. If AI systems are trained on historical data that reflects past biases, they can perpetuate and even amplify those biases.
This is not the first time Meta has faced scrutiny over its use of AI in employment. In 2022, the company settled a lawsuit brought by the US Department of Justice that accused it of using AI-driven ad targeting to discriminate against older workers. That case involved Meta’s advertising platform, which allowed employers to exclude users over 40 from seeing job ads. The settlement required Meta to pay over $115,000 in penalties and change its ad delivery systems. More recently, in 2024, a group of Meta contractors filed a class action alleging that the company’s productivity algorithms unfairly penalised them for taking breaks, leading to unfair terminations. That case is still pending.
The Technology Behind the Suit
At the heart of the complaint is Metamate, an internal large language model assistant that Meta began rolling out to employees in 2023. Metamate can summarize meetings, answer questions, and generate code, but the plaintiffs allege it was also used to generate “productivity scores” based on how often employees used AI tools like tokens and prompt submissions. The lawsuit claims that the more an employee engaged with Metamate, the higher their score, making those on leave appear less productive.
Additionally, the “second brain” agents mentioned in the suit are AI systems trained on individual employees’ work habits, writing style, and project history. These agents were ostensibly designed to help workers recall information, but the complaint says they also fed into performance rankings. Employees who were absent had fewer interactions with these tools, leading to lower data density and, in turn, lower algorithmic scores.
Keystroke and activity monitoring—capturing every mouse movement, click, and typed character—is common in customer service and call center industries but has been controversial in knowledge-work settings. Meta’s monitoring program, allegedly rolled out in early 2026 without clear consent, recorded not only keystrokes but also browser history, email content, and location data from company devices. The plaintiffs argue that this surveillance created a chilling effect and that its use in layoff decisions violates both privacy expectations and employment law.
Impact on the Tech Industry
The lawsuit could set a precedent for how Big Tech handles workforce reductions in an era of pervasive AI. Since late 2022, major companies including Google, Amazon, Microsoft, and Meta have collectively laid off hundreds of thousands of workers, often citing the need to streamline operations and invest in AI. As these companies rely more on algorithmic tools to manage performance, the risk of legal challenge increases. If the court allows class certification, thousands of former Meta employees could potentially join the case.
Employment lawyers are watching closely. “This is a landmark case because it directly challenges the black box of AI decision-making in the workplace,” said Sarah Hennessey, a partner at a San Francisco-based labor law firm not involved in the case. “We’re moving from asking whether AI is fair to asking whether it violates specific legal protections. The plaintiffs here have a strong narrative: the same tools that were supposed to enhance productivity were used to cut people who were sick or caring for family.”
Legal experts also note that the case underscores the tension between efficiency and equity. Meta’s argument that human managers made the final call may not shield it from liability if the AI systems essentially predetermined the outcomes. Under the ADA, employers cannot use neutral policies that have a disparate impact on disabled employees unless they can show the policy is job-related and consistent with business necessity. The plaintiffs argue that using AI tokens and activity data fails that test because it penalizes legitimate absences.
Future of AI in Employment Decisions
Regulatory action is accelerating. In addition to California and New York City laws, the European Union’s AI Act, which entered into force in 2024, classifies employment-related AI systems as high-risk, requiring conformity assessments, human oversight, and transparency. Meta, like many tech giants, has operations in Europe and may face additional scrutiny there. The US Congress, meanwhile, has yet to pass comprehensive AI legislation, though several bills have been introduced that would require bias testing and impact assessments for automated employment tools.
The outcome of this lawsuit could influence how quickly those regulations are implemented. If Meta loses and is forced to pay substantial damages or implement changes, other companies may rush to audit their own AI systems. Conversely, if the court sides with Meta, it could embolden employers to expand algorithmic monitoring and scoring without fear of legal repercussions.
For now, the 26 plaintiffs remain anonymous, but their stories are emblematic of a broader anxiety among tech workers: that the very machines they help build are now being used to judge them. As the July 22 termination date approaches, the urgency of their request for an injunction grows. If granted, it would temporarily halt their layoffs while courts determine whether AI-driven downsizing can coexist with the legal protections that workers have fought decades to secure.
The case is Doe v. Meta Platforms Inc., and its next major hearing is expected in August. Whatever the outcome, it will be a crucial test of how far the law can stretch to protect workers in an age of algorithmic management.