Meta sued for using discriminatory AI in mass layoffs

What happened?

In May 2026, Meta laid off 8,000 employees (10% of its workforce) as part of a restructuring to redirect funds toward massive investments in artificial intelligence. However, a class-action lawsuit filed on July 14 in the Northern District of California by 26 anonymous former employees alleges that the selection of those laid off was not made by human supervisors but by a set of internal AI systems that discriminated against those who had taken medical, parental, or maternity leave. This case, reported by Ece Yildirim at Gizmodo, marks a milestone in litigation over algorithmic bias in human resources.

The AI systems involved

According to the complaint, Meta used a constellation of algorithmic tools to score and rank employees, including:

• Metamate: an assistant based on a large language model trained on internal communications and documents.
• Algorithmic productivity scores: based on data such as keystrokes, browsing history, and email activity.
• AI-assisted performance review tools.
• Internal resource consumption logs.

The lawsuit argues that these systems systematically penalized employees who had been absent for legally protected reasons. A telling case: one plaintiff was selected for layoff when she was two days away from giving birth; another, on disability leave related to pregnancy, was the only one from her team on the list. The complaint notes that Meta did not use the informed judgment of managers but relied on opaque and potentially biased systems. This pattern suggests that the algorithms learned to associate absences with low performance, without considering legal protections.

Why is this important?

This case sets a critical precedent for the use of AI in employment decisions. Unlike ordinary labor disputes, it questions the legality of delegating to algorithms decisions that may violate fundamental rights, such as protection against discrimination based on pregnancy or disability. The lawsuit alleges that Meta did not use the informed judgment of managers but relied on opaque and potentially biased systems. Moreover, it occurs in a context where AI regulation is advancing slowly: the European Union approved its AI Act in 2024, but its implementation is not yet complete; in the United States, there is no specific federal law, although the FTC has shown interest in algorithmic fairness. This regulatory vacuum makes cases like this crucial for defining responsibilities.

“Meta did not compile the layoff list through the considered judgment of managers who knew the work,” the complaint states. “Instead, it turned to a constellation of internal artificial intelligence systems.”

Potential consequences

If the lawsuit succeeds, it could have far-reaching implications:

• For Meta: potential financial penalties, which could amount to millions of dollars in compensation and punitive damages. Additionally, it would be forced to audit its algorithms and implement changes in its human resources practices, such as mandatory human oversight in layoff decisions.
• For the tech industry: it could establish a standard of accountability for the use of AI in hiring and firing processes, forcing companies like Google, Amazon, or Microsoft to review their own tools. This could slow down the adoption of AI in HR or incentivize the creation of internal ethics committees.
• For workers: it reinforces the need for transparency and human oversight in automated decisions that affect labor rights. It could also boost demand for laws like the Algorithmic Accountability Act in the U.S., which would require companies to assess the bias of their systems.

Context and comparisons

This is not the first time a major tech company has faced allegations of algorithmic bias. Amazon had to scrap an AI hiring system that discriminated against women in 2018. However, this case is unique because it applies to mass layoffs and the severity of the consequences for those affected. Moreover, it occurs at a time when AI regulation is advancing slowly: the EU approved its AI Act in 2024, but its implementation is not yet complete. In the U.S., there is no specific federal law, although the FTC has shown interest in algorithmic fairness. Unlike the Amazon case, where the system was scrapped before being used at scale, here the AI has already caused real harm to 8,000 people. It also differs from other algorithmic discrimination lawsuits, such as iTutorGroup in 2022 (which discriminated by age in hiring), due to the number of affected individuals and the massive nature of the layoffs.

What readers should know

This case underscores that AI is not neutral: it reproduces and amplifies biases present in the data it is trained on. Companies using AI for employment decisions must implement rigorous audits, ensure human oversight, and comply with anti-discrimination laws. For workers, it is crucial to know what data is collected and how it is used to evaluate their performance. The lawsuit against Meta is a reminder that technology should serve people, not harm them. Moreover, it raises urgent questions: should employees have the right to challenge automated decisions? What transparency should companies offer about their algorithms? The answers to these questions will define the future of work in the age of AI.