A new high-profile scandal is brewing in the industry, casting a shadow on the ethics of using artificial intelligence in human resources management. 26 current and former Meta employees have filed a class-action lawsuit in the U.S. District Court for the Northern District of California, alleging that the company used AI systems to improperly select candidates for layoffs. The plaintiffs, who are proceeding anonymously due to fear of retaliation, claim that the algorithms disproportionately "weeded out" people who were on maternity, family, or medical leave, as well as employees with disabilities.
The case concerns a large-scale restructuring in which Meta plans to cut about 10% of its workforce — approximately 8,000 people. Layoff notices were sent out on May 20, with actual termination of employment for most scheduled for July 22. The plaintiffs' key argument: the list for cuts was not formed based on balanced decisions by managers, but on a combination of internal AI tools.
How did the "digital guillotine" work?
According to the plaintiffs, Meta deployed a whole arsenal of algorithms: the corporate assistant Metamate, so-called "second brain" agents, data on AI token usage, automated performance ratings, and information on actions on corporate devices. These metrics accounted for the volume of work completed, activity, number of code commits, and results of annual reviews. The problem is that employees on parental leave, sick leave, or maternity leave were physically unable to accumulate these indicators on par with their working colleagues.
The plaintiffs emphasize: Meta did not adjust the ratings to account for legitimate periods of absence and did not conduct an individual review of the results. One plaintiff — a scientist who was on approved leave before giving birth — received a layoff notice two days before her child was born. Another employee stated that his rating was lowered due to a break related to an injury. A manager on medical leave was included on the layoff list just 16 days into their absence.
Meta's position and the plaintiffs' demands
A Meta representative called the claims unfounded, stating that all personnel decisions are made by people, not AI. However, the plaintiffs insist otherwise and cite laws prohibiting discrimination against pregnant people and individuals with disabilities. They demand a temporary halt to the layoffs, preservation of salaries, health insurance, and stock rights until the proceedings conclude. They are also seeking an independent audit of the AI tools used. Lawyers warn: once the layoffs are completed, the consequences will become irreversible — employees will lose insurance during pregnancy, postpartum recovery, and medical treatment, and in some cases, their immigration status.
This lawsuit is just the tip of the iceberg. It was previously reported that Meta installed a Model Capability Initiative system on corporate devices, capable of recording keystrokes, mouse movements, and screen content to train AI agents. After widespread discontent (a petition was signed by over 1,600 employees), Mark Zuckerberg suspended the program, but the resentment remains.
My analysis: This case is a warning sign for the entire industry. Tech giants are increasingly trying to automate "uncomfortable" decisions, such as mass layoffs, shifting responsibility onto algorithms. However, a mathematical model, devoid of empathy and understanding of human circumstances, inevitably becomes a tool of discrimination. If the court finds Meta guilty, it will set a precedent that forces a global reassessment of approaches to implementing AI in HR.