What happens when your HR department is AI-powered and handles admin for workers? Well, because of that, AI now finds itself at the core of a legal battle over one of the most sensitive jobs in any business: deciding who keeps their job and who does not.
Dozens of Meta employees have taken the company to court, claiming AI systems helped identify workers for redundancy. According to a lawsuit filed in the federal court for the Northern District of California, the systems penalised people who had taken maternity leave, medical leave or disability accommodation.
The case has turned attention to a difficult issue for employers. If software helps score, rank and sort employees, who is really making the decision?
What Do The Workers Claim Happened?
The lawsuit covers about 8,000 job losses announced earlier this year at Meta, the owner of Facebook, Instagram and WhatsApp.
According to the complaint, 26 employees allege that Meta used what it calls a “constellation of internal artificial intelligence systems” to help identify people for redundancy. The filing says the company relied on AI performance ratings, keystroke monitoring, activity data and productivity measures when deciding who should appear on the redundancy list.
The complaint says, “Meta did not assemble the termination list through the considered judgment of managers who knew the work.” It continues, “Instead, Meta used artificial intelligence systems to score, rank and select employees for inclusion on the list.”
The workers are asking the court to stop the redundancies from being finalised while the legal case continues. They are also asking for reinstatement, back pay, lost equity, employee benefits and other damages if they succeed.
More from Artificial Intelligence
- Halifax Startup Launches The UK’s First AI-Powered Instant Motorhome Valuation Platform, Motorhome Pig
- AI At A Crossroads: Why Trust, Talent, Cost And Delivery Will Define The Next Phase Of AI
- Could US Businesses Face Federal Deception Penalties For Using Biased AI Tools?
- Could AI Voice Licensing Become The Next Celebrity Revenue Stream?
- Are US Businesses Wasting Money On AI Supply Chain Tools That Don’t Deliver?
- Is Meta Muse A Creative Breakthrough Or A Privacy Concern?
- The UN Wants Global AI Guardrails, But What Would That Mean For Startups?
- Would You Trust AI To Tell You Where To Seek Medical Help?
How Could Leave Affect An AI Score?
A major section of the complaint deals with protected leave because according to the lawsuit, the AI systems collected information about performance rankings, productivity and workplace activity. Employees on maternity leave or medical leave naturally produced less activity during those periods. Workers with disabilities could also record lower productivity measures.
The complaint says, “The result was that employees who took protected leaves were disproportionately selected for layoff, based on scoring that not only failed to account for their protected leaves, but in effect penalized the employees for exercising their legal rights to these leaves.”
The filing gives personal examples from the group bringing the claim. One scientist says she received notice that she had lost her job just two days before giving birth during approved pre birth pregnancy leave. Another engineer says an injury and time away from work resulted in a lower rating. A manager says he lost his job only 16 days after beginning medical leave.
Reuters also reported that the lawsuit claims Meta used opaque productivity measures and large language model token usage when selecting workers for redundancy. According to Reuters, the 26 plaintiffs include engineers, researchers, designers, managers and a director who had direct access to the AI platform.
What Does Meta Say?
Meta rejects every allegation made in the lawsuit after a company spokesperson told The Guardian, “These claims lack merit and are not based on facts. Workforce management and organizational decisions were and are made by people, not AI.”
The lawsuit also examines Meta’s employee monitoring programme introduced earlier this year. According to the complaint, the system collected keystrokes, mouse activity, browser history, emails, messages and location data from company devices.
According to The Information, Meta chief executive Mark Zuckerberg explained the thinking behind the programme during an internal meeting. He said, “The AI models learn from watching really smart people do things. The average intelligence of the people who are at this company is significantly higher than the average set of people that you can get to do tasks.”
The complaint also claims employees were informed through what it calls a low visibility internal post written by an engineer instead of a senior leader. It says many employees did not receive a consent prompt and initially had no option to opt out.
The legal case has only just begun, and the court has not ruled on the allegations. The outcome could influence how employers use artificial intelligence when making decisions about their workforce.
