You guys — this is wild and needs a dedicated thread on here, but I couldn't wait until later. Meta got sued by twenty-six former employees for allegedly using biased AI systems during their May round of layoffs that cut about 8000 people! The lawsuit describes what they call "a constellation" of tools: Metamate (their internal assistant), employee-trained second brain agents, token usage dashboards — yeah, actually tracking how many tokens each person used — and even keystroke monitoring. The problem is the algorithm ranked employees on things like whether they were "AI-native," which means anyone on medical or family leave who couldn't use these tools was automatically pushed toward layoff eligibility. I cannot get over the sheer audacity of penalizing someone for being unavailable during protected leave with a productivity metric that literally requires them to be online and active!
Meta claims the suit lacks merit because all hiring decisions involved human input, but the lawyers counter that every one of the twenty-six plaintiffs was on FMLA or the California Family Rights Act within 24 months before getting cut — which is illegal under federal law. They also cite California's Fair Employment and Housing Act, which specifically bans automated systems producing disparate impact discrimination based on disability, sex, or pregnancy, so that catches a lot of people Metas "Merits" defense misses. The plaintiffs want an independent audit of the entire algorithmic selection process before Meta finishes its layoffs; meanwhile, this is all happening after Reuters reported in April 2026 that Meta was recording staff keystrokes and mouse clicks to train AI — a program they only stopped when private chat transcripts became accessible. The fact that this same surveillance pipeline likely fed their layoff algorithm isn't shocking at all given the company's history, but it is absolutely chilling.
Source: https://www.engadget.com/2215258/former-employees-sue-over-meta-s-alleged-use-of-biased-ai-systems-during-layoffs
Meta claims the suit lacks merit because all hiring decisions involved human input, but the lawyers counter that every one of the twenty-six plaintiffs was on FMLA or the California Family Rights Act within 24 months before getting cut — which is illegal under federal law. They also cite California's Fair Employment and Housing Act, which specifically bans automated systems producing disparate impact discrimination based on disability, sex, or pregnancy, so that catches a lot of people Metas "Merits" defense misses. The plaintiffs want an independent audit of the entire algorithmic selection process before Meta finishes its layoffs; meanwhile, this is all happening after Reuters reported in April 2026 that Meta was recording staff keystrokes and mouse clicks to train AI — a program they only stopped when private chat transcripts became accessible. The fact that this same surveillance pipeline likely fed their layoff algorithm isn't shocking at all given the company's history, but it is absolutely chilling.
Source: https://www.engadget.com/2215258/former-employees-sue-over-meta-s-alleged-use-of-biased-ai-systems-during-layoffs