Key Takeaways
- A group of 26 Meta employees has sued the company, claiming that internal AI systems were used to select workers for layoffs after they took protected leave or requested disability accommodations.
- The lawsuit alleges that Meta’s “constellation of internal artificial intelligence systems”—including performance‑rating models and keystroke‑activity monitors—scored and ranked employees, disproportionately penalizing those on medical, maternity, or disability leave.
- Plaintiffs seek a preliminary injunction to halt the layoffs, reinstatement, back pay, lost equity, benefits, and other damages, and request an independent audit of the AI tools used in the decision‑making process.
- Meta denies the allegations, stating that workforce decisions are made by people, not AI, and that the employee‑monitoring program was paused in June after widespread privacy concerns.
- The case highlights growing legal and regulatory scrutiny of AI‑driven workplace decisions, with states such as California, Colorado, and Illinois enacting laws to protect workers from biased automated systems.
Overview of the Lawsuit
On Monday, a 71‑page complaint was filed in the U.S. District Court for the Northern District of California by 26 current Meta employees. The suit names Meta Platforms, Inc.—the parent company of Facebook, Instagram, and WhatsApp—as the defendant. The plaintiffs contend that the company’s recent workforce reduction, which eliminated roughly 8,000 positions earlier this year, was not the result of managerial judgment but rather the output of artificial‑intelligence tools that systematically identified employees for termination. They are asking the court for a preliminary ruling to stop Meta from finalizing the layoffs while the case proceeds, as well as monetary relief that could include reinstatement, back pay, forfeited equity, lost benefits, and other damages.
Allegations of AI‑Driven Targeting
The core of the plaintiffs’ argument is that Meta deployed a “constellation of internal artificial intelligence systems” to score, rank, and select workers for the layoff list. According to the complaint, these systems ingested data from performance ratings, productivity metrics, and extensive monitoring of employee activity—including keystrokes, mouse movements, browser history, emails, messages, and location data on company devices. The lawsuit asserts that the AI models used this information to generate a layoff score, and that the resulting rankings disproportionately flagged employees who had taken protected leave or requested disability accommodations. In essence, the plaintiffs claim the AI punished workers for exercising legally protected rights.
Impact on Employees on Leave
Several concrete examples illustrate how the alleged AI bias manifested. One plaintiff, a scientist who was on approved pre‑birth pregnancy leave, received notice of her layoff just two days before she gave birth. Another plaintiff, an engineer, said his performance rating was lowered because he had taken time off to recover from an injury. A third plaintiff, a manager on medical leave, reported being terminated only 16 days into his approved leave period. The complaint argues that because the AI systems did not account for periods when employees were not actively working—such as during maternity, medical, or disability leave—their scores were artificially depressed, making them more likely to be selected for termination. This, the plaintiffs contend, amounts to discrimination based on protected characteristics.
Meta’s Response and Program Details
A Meta spokesperson rejected the allegations, stating in an email to the Guardian that “these claims lack merit and are not based on facts” and that workforce decisions are made by people, not AI. Nevertheless, the lawsuit notes that Meta introduced an AI‑employee‑monitoring program earlier this year, designed to capture a wide range of behavioral data from company devices. CEO Mark Zuckerberg reportedly described the initiative in an internal meeting as a way to train AI models by observing “really smart people” perform tasks, claiming the average intelligence at Meta exceeds that of the general population. The program was rolled out without formal employee consent; workers learned of it through a low‑visibility internal post by an engineer, and on some teams no acknowledgment or opt‑out mechanism existed initially.
Employee Backlash and Program Pause
Within months, the monitoring program provoked significant employee pushback. More than 1,600 staff members signed a petition alleging the program violated their privacy, prompting Zuckerberg to announce a pause of the initiative in June. The lawsuit cites this backlash as evidence that the program was controversial and that Meta proceeded with limited transparency. The plaintiffs argue that the lack of clear communication and consent undermines any claim that the AI tools were used responsibly or fairly.
Requests for Judicial Intervention
The plaintiffs’ lawyers have asked the court to order an immediate, independent audit of Meta’s AI tools to determine how the layoff selections were made and why the 26 workers on leave or disability accommodation were disproportionately affected. They contend that Meta deliberately concealed the mechanics of its selection process from employees, violating principles of transparency and fairness. Additionally, because the employees remain on Meta’s payroll until their terminations are set to begin on July 22, the attorneys seek anonymity for the plaintiffs and a court order preserving their employment status while arbitration proceeds, warning that any final separation would cause irreversible harms such as loss of employer‑subsidized health coverage during pregnancy or medical treatment, forfeited unvested equity, and potential immigration consequences.
Broader Context: AI in the Workplace and Emerging Regulations
The Meta case sits within a broader trend of increasing scrutiny over AI‑driven employment decisions. Workers across industries have raised concerns about bias, privacy, and erosion of trust when algorithms influence hiring, promotions, performance evaluations, and layoffs. In response, several states have enacted legislation aimed at curbing discriminatory outcomes from automated decision systems. California, Colorado, and Illinois have each passed laws or regulations in the last couple of years that require employers to disclose the use of AI in employment decisions, conduct bias audits, and provide avenues for redress. The lawsuit against Meta may serve as a test case for how these new rules apply to large tech firms that rely heavily on internal AI analytics for workforce management.
Conclusion
The lawsuit against Meta underscores the tension between leveraging advanced AI for operational efficiency and safeguarding employee rights. If the plaintiffs succeed, the case could compel greater transparency, stricter oversight, and potentially significant financial remedies for workers allegedly harmed by algorithmic layoff decisions. Conversely, a ruling in Meta’s favor might reinforce the notion that AI tools, when properly governed, can be used without violating anti‑discrimination laws. Either outcome will likely influence how other companies design, deploy, and monitor AI systems in the workplace, and how regulators shape the legal landscape surrounding automated employment decisions.