A group of 26 current and former Meta employees has sued the social media company, alleging artificial intelligence-linked metrics used around its latest layoffs put workers on protected leave or with disabilities at a disadvantage.
The legal complaint, filed on Monday in the US District Court for the Northern District of California, centres on Meta's May layoffs, which affected 10% of the company's workforce, according to CNBC.
The workers allege Meta's internal AI systems failed to properly account for approved absences when assessing employees for job cuts.
Meta has rejected the claims.
Workers question how AI-era performance is measured
At the centre of the lawsuit is a growing workplace concern: what happens when employee performance metrics cannot capture why someone produced less?
Lawyers representing the 26 unnamed workers allege Meta relied on a collection of internal AI systems drawing on performance ratings, calibration scores and productivity measures. The complaint also points to “AI-native” ratings and AI-token consumption.
According to the filing, employees on protected medical or family leave could not accumulate some of these measures in the same way as colleagues who were actively working. Employees whose output was affected by a disability could face a similar problem.
“Those tools draw on inputs—performance ratings, calibration scores, productivity and output metrics, ‘AI-native’ ratings, and AI-token consumption—that, by design, cannot be accumulated by an employee who is on protected medical or family leave, or whose output is reduced by a disability,” lawyers for the workers wrote in the complaint, as quoted by CNBC.
The lawsuit alleges Meta violated protected-leave laws and discrimination laws related to pregnancy and disability, among other areas. The plaintiffs said they intend to pursue their individual claims through arbitration.
Courthouse News Service first reported on the lawsuit.
What the Meta workers allege
The complaint puts several elements of Meta's layoff process under scrutiny:
- 26 current and former employees are involved in the legal action.
- The workers were among employees affected by Meta's May layoff round, which cut 10% of its workforce, according to the complaint cited by CNBC.
- The plaintiffs allege internal AI systems failed to account for approved absences.
- The filing questions the use of performance, productivity, AI-native and AI-token consumption metrics.
- The workers allege such measures could disadvantage employees on medical or family leave and people whose output was reduced by a disability.
- The plaintiffs are seeking a preliminary injunction and an independent audit of the allegedly algorithm-assisted selection process.
The workers want the court to maintain the status quo of their employment while the selection process is independently audited and their claims move through arbitration.
Meta says people, not AI, made workforce decisions
Meta disputes the workers' account.
A company spokesperson told CNBC the claims “lack merit and are not based on facts”.
“Workforce management and organizational decisions were and are made by people, not AI,” the spokesperson said.
The response creates a key point of dispute in the case. The plaintiffs allege algorithmically assisted systems influenced the selection process. Meta maintains people made its workforce and organisational decisions.
The court proceedings could put greater attention on the distinction between AI making an employment decision and AI-generated metrics informing a human decision.
AI's role in workplace decisions faces closer scrutiny
The Meta lawsuit arrives as the use of AI in employment systems faces growing legal attention in the US.
Nearly a month earlier, a federal judge in California ruled that Workday must face claims concerning its AI-powered job screening services, CNBC reported. The lawsuit alleges the technology violated state and federal discrimination laws.
Workday has denied the allegations.
The company said its recruiting software does not make hiring decisions “in California or anywhere else” and assesses job qualifications rather than protected characteristics such as race, age or disability.
Workday also said it tests its products through its Responsible AI programme to check whether its tools harm protected groups.
The Meta case concerns layoffs rather than recruitment, but both disputes raise a broader question for employers adopting AI-linked systems: how should organisations test whether a seemingly neutral metric produces unequal consequences for certain workers?
For employees, the concern is equally immediate. A productivity score can measure output. AI usage data can measure interaction with a tool. Neither automatically explains an absence, a disability or the circumstances behind a period of lower activity.
As AI becomes more deeply embedded in performance and workforce systems, the debate may increasingly shift from whether a human signed off on a decision to what data shaped the decision in the first place.
The allegations against Meta remain contested. But the lawsuit puts a new layer of the AI-at-work debate under the spotlight: when algorithms help measure employees, employers may face tougher questions about what those systems fail to see.