These days, corporations aren’t just using AI as an excuse to justify layoffs, but are increasingly turning to their chatbots for advice deciding who to fire.

As Reuters reports, 26 Meta employees filed a lawsuit accusing the company of using an internal AI platform to select targets for the thousands of jobs cut earlier this year. Per the suit, Meta used its AI to cull workers based on opaque productivity metrics and large language model (LLM) token usage, a move which disproportionately targeted people with disabilities, and those who had taken maternity or medical leave — literally adding insult to injury.

The 26 plaintiffs, whose names have been redacted from the suit, include engineers, managers, researchers, and designers, as well as a director who had direct access to the AI platform.

Nearly half of the plaintiffs say that their termination was the result of taking either maternity or parental leave, protected activities which Meta’s AI-integrated human resources platform “Checkpoint” allegedly failed to take into consideration. Others claim to have been targeted after taking medical leave for their own disabilities, bereavement leave, or to care for family members.

In some cases, plaintiffs allege to have been specifically warned by Meta leadership not to take time off. The anonymous director, for example, says he was “discouraged and deterred from taking FMLA [Family and Medical Leave Act] leave that his treating provider had recommended, out of a well-founded and objective concern from a manager that such a leave would be held against him in the [reduction in force].”

In the suit, the director corroborates the claim that Meta’s Checkpoint used protected leaves of absence against workers, arguing that the platform “failed to distinguish employees’ leave from lack of engagement.”

The current complaint seeks to block Meta from laying off the 26 plaintiffs, who will each have to file an independent suit owing to legal agreements with the company, Reuters reports.

In a statement, a Meta spokesperson disputed the basic facts of the case.

“Workforce management and organizational decisions were and are made by ​people, not AI,” the spokesperson told Reuters.

More on Meta: Meta Caught Running Ads for Child Abuse