Twenty-six current and former Meta employees have filed a federal lawsuit in Oakland, California, accusing the company of using artificial intelligence systems that unfairly targeted workers with disabilities or those who had taken medical leave for protected reasons for layoffs. The suit, the first of its kind against a major US technology company, challenges Meta's use of algorithmic decision-making in workforce reductions, alleging violations of the Americans with Disabilities Act, the Family and Medical Leave Act, the Pregnancy Discrimination Act, and various state-level protections.
The plaintiffs, who filed anonymously share one common trait: each had taken, requested, or been approved for protected leave within the 24 months preceding the layoffs. The layoffs, which began on May 20 and affected roughly 8,000 positions, were part of Meta's broader cost-cutting measures. The complaint alleges that Meta employed a constellation of internal AI systems to score, rank, and select employees for termination, including Metamate (a large language model assistant), employee-trained “second brain” agents, keystroke and activity monitoring data, AI token usage dashboards, and algorithmically assisted performance rankings.
According to the 71-page complaint, these systems disproportionately disadvantaged employees who had taken time away from work because they had fewer data points to measure. The plaintiffs argue that the reliance on output metrics inherently biases against workers with health conditions or caregiving responsibilities, effectively punishing them for using legally protected leave.
Meta has faced intense scrutiny over its use of AI in human resources decisions, a growing area of concern among regulators worldwide. The company previously announced a major restructuring in which AI was used to identify roles for elimination, but it denied that AI alone made the final decisions. In a statement, a Meta spokesperson said the claims lack merit, emphasizing that “workforce management and organisational decisions were and are made by people, not AI.” The spokesperson also noted that the company employed human reviewers to validate the AI-generated lists, a common practice among tech firms using algorithmic tools in layoffs.
The lawsuit also details a monitoring program that Meta allegedly deployed earlier in 2023, which 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 this monitoring program was used to build AI tools, including by a new engineering organization to which employees were reassigned on a mandatory basis.
This case arrives amid a broader debate about the role of artificial intelligence in employment decisions. Courts and regulators around the world are grappling with how existing labour laws apply 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 has issued guidance stating that AI systems used in hiring or firing must comply with anti-discrimination laws, but enforcement remains uneven. Some states, including California and New York City, have recently passed laws requiring companies to audit their AI tools for bias, and the plaintiffs in this case allege Meta failed to test its AI systems for such bias, violating those regulations.
The legal challenge is also significant because it comes from within the company—employees who experienced the AI's impact firsthand. While Meta's workforce has been reduced by over 20,000 jobs since late 2022 (including the 11,000 layoffs in November 2022 and the subsequent 8,000 in May 2023), this lawsuit focuses specifically on the algorithmically guided selection for the May layoffs. 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. Judge William Orrick in the Northern District of California has been assigned to the case.
The allegations raise fundamental questions about the fairness of using data-driven performance metrics in human resources. In an environment where employees are constantly monitored and measured, the use of AI to automate termination decisions can inadvertently encode biases present in historical data. For instance, if employees who take medical leave are historically less productive in terms of measurable output, the algorithm may learn to associate leave with lower value, regardless of the legal protections for that leave. Critics argue that such systems lack the nuance to account for circumstances like illness, pregnancy, or caregiving, and that they effectively penalize workers for exercising their legal rights.
Meta is not the only company facing such scrutiny. In recent years, Amazon was criticized for using an AI hiring system that penalized women by favoring resumes with male-specific language, and the company eventually scrapped that tool. Other tech giants like Microsoft and Google have also faced internal pressure to disclose how AI is used in employee evaluations. However, this lawsuit is the first to directly challenge the use of AI in selecting employees for mass layoffs, setting a potential precedent for future litigation.
The case also highlights the tension between corporate cost-cutting and employee rights. Meta, like many tech companies, has been aggressively reducing its workforce to improve profitability amid a challenging macroeconomic environment. The company's stock has rebounded since the layoff announcements, but the human cost has been substantial. Employees affected by the layoffs have reported feeling that the process was opaque and arbitrary, with many learning their fate through automated emails or portal updates. This lawsuit seeks to prove that the process was not just arbitrary but discriminatory.
From a legal perspective, the plaintiffs face an uphill battle. To prove discrimination under the ADA or FMLA, they must show that their protected activity (taking leave) was a motivating factor in the adverse employment action. If Meta can demonstrate that it had legitimate, non-discriminatory reasons for the layoffs—such as overall cost savings or strategic realignment—the plaintiffs may struggle to establish causation. However, the detailed allegations about the AI system's design and its reliance on historic productivity metrics could bolster their case by showing that the system was inherently discriminatory.
Another key issue is whether Meta can be held liable for the outputs of its self-trained AI systems. The company has argued that it used human oversight, but the complaint contends that the human reviewers were largely rubber-stamping the AI's recommendations, and that the design of the system—including the use of token usage dashboards and keystroke monitoring—was to blame. If the court finds that Meta failed to take reasonable steps to prevent discriminatory outcomes, it could face substantial damages, including back pay, front pay, and punitive damages.
This lawsuit also occurs against the backdrop of increasing regulation of AI in employment. New York City's Local Law 144, which took effect in 2023, requires employers to conduct bias audits of automated employment decision tools and publish the results. California's similar law, AB-53, goes into effect in 2024. Meta, which has a large presence in both locations, must comply with these rules. The plaintiffs argue that Meta's failure to conduct such audits before using the AI systems in the layoffs constitutes a direct violation of these laws.
The case has attracted attention from civil rights groups and labour organizations. The American Civil Liberties Union has filed amicus briefs in similar AI bias cases, and the National Employment Law Project has called for greater oversight of algorithmic decision-making in the workplace. Some legal experts believe that if the Meta case succeeds, it could lead to sweeping changes in how companies use AI for workforce management, forcing them to implement more rigorous testing and transparency measures.
In the immediate term, the plaintiffs are seeking to stop their terminations through a preliminary injunction. That request will be heard by Judge Orrick, who will weigh the likelihood of success on the merits against the potential harm to Meta. If the injunction is granted, it would effectively pause the layoffs for the named plaintiffs and set a precedent requiring the company to reconsider its selection process for all affected employees.
The Meta lawsuit is a harbinger of more litigation to come. As companies increasingly rely on AI to manage their workforces, the legal system is being asked to adapt decades-old anti-discrimination laws to new technological realities. Whether the plaintiffs succeed or fail, the case will almost certainly shape the future of AI in human resources, forcing companies to reexamine how they design and deploy these systems.
For now, the 26 employees wait for their day in court, hoping to prove that a machine should not have the final say over their livelihoods. Their case raises profound questions about accountability, fairness, and the intersection of artificial intelligence with fundamental worker protections. As the July 22 deadline for their terminations approaches, the outcome of this lawsuit could send shockwaves through Silicon Valley and beyond.