Meta’s tool, called Metamate, reportedly scored, ranked and selected employees for a sweeping staff reduction that took place in May across the company. The federal lawsuit, filed in Oakland, California, argues that staff on pregnancy, medical or family leave could not log the code commits, output volume and AI-tool usage that Metamate tracked. Those metrics sat at the core of how the system evaluated people and, according to the claim, who ended up losing their jobs.
Meta, which also owns Instagram and WhatsApp, faces allegations that this setup allowed legally protected absences to be used against employees. Claimants say every one of the 26 affected workers had taken protected leave in the two years before the cuts, leaving a clear pattern in who the AI flagged.
Because they were away, they generated fewer of the signals Metamate valued most, from code pushes to interaction with company AI tools. Legal and workplace experts say such designs can bake bias into algorithms without any human manager consciously deciding to discriminate.
Internal performance systems that hitch rankings to activity logs can disadvantage carers, people recovering from illness and anyone who steps back temporarily. Similar tensions around AI-driven “efficiency” are emerging in Australia, where major technology employers are also cutting staff as they embrace automation.
Atlassian, Afterpay owner Block and WiseTech Global have each axed significant numbers of roles while publicly leaning into AI to reshape their businesses. The Meta lawsuit highlights how performance analytics can collide with employment law when tools are trained on narrow productivity signals.
