A lawsuit against Meta Platforms is drawing attention to a growing workplace challenge. Employees who believe artificial intelligence influenced decisions about their jobs often struggle to prove it.
According to reporting by Reuters, the case alleges Meta used discriminatory AI tools when selecting employees for layoffs. The lawsuit has also exposed the legal and practical barriers workers face when attempting to challenge the use of AI in employment decisions.
Meta has denied the allegations, maintaining that people, not AI, made all layoff decisions.
Workers face an uphill battle
Legal experts told Reuters that employees frequently have little visibility into how AI systems are used inside organisations.
The challenge becomes even greater when workers are required to resolve disputes through arbitration rather than in court.
According to the report, arbitration agreements often prevent employees from:
- Filing class-action lawsuits.
- Presenting their case before a jury.
- Seeking damages through public court proceedings.
- Accessing evidence that could reveal how employment decisions were made.
Because arbitration proceedings remain confidential, evidence uncovered in one case may never become available to other employees facing similar circumstances.
Judge points to lack of access
Reuters reported that US District Judge William Orrick, in a ruling issued last week, highlighted one of the central difficulties confronting the plaintiffs.
The judge declined to stop Meta from completing the termination of 26 employees, noting that the workers "were not in the rooms where it happened."
According to Reuters, the employees allege they were selected for layoffs because of disabilities or because they had taken medical leave. However, they lack direct evidence showing how the company's employment decisions were reached.
Lawsuit details AI allegations
According to the complaint cited by Reuters, the plaintiffs allege Meta relied on AI-assisted systems when determining which roles to eliminate.
The lawsuit claims the company used:
- Productivity measurements.
- AI token usage, a metric tracking employees' use of AI tools.
- An internal large language model assistant known as Metamate.
- An employee-developed knowledge system described as a "second brain".
- Productivity scores generated from keystrokes, screen activity, emails and browser history.
The plaintiffs claim these systems influenced layoff decisions.
Meta rejects the claims
In court filings and public statements cited by Reuters, Meta said all decisions relating to the approximately 8,000 layoffs announced earlier this year were made by people.
The company denied using AI to identify employees for termination or to conduct performance appraisals.
A Meta spokesperson told Reuters the company had no additional comment beyond its earlier statements.
Arbitration limits broader legal challenges
According to Reuters, employment lawyers believe the Meta case also explains why relatively few lawsuits have emerged over workplace AI, despite its increasing adoption.
Christine Webber, co-chair of the civil rights and employment practice at Cohen Milstein Sellers & Toll, told Reuters that even if an AI system were shown to produce discriminatory outcomes, confidential arbitration makes it difficult for employees to share evidence with others.
Reuters noted that one of the few other prominent workplace AI cases involves Workday, which has been accused of using AI that discriminated against job applicants based on race, age and disability. Workday has denied those allegations.
Unlike Meta's employees, applicants in the Workday case are not bound by employment arbitration agreements with the software provider, making public litigation possible.
Debate over AI transparency continues
As organisations expand the use of AI across workplace processes, the Meta lawsuit has brought renewed attention to questions of transparency, accountability and employee rights.
The outcome of the case could influence future discussions around how companies disclose the use of AI in employment decisions and how workers can challenge decisions they believe were shaped by automated systems.