July 20, 2026
Judge Won’t Block Meta AI Layoffs, But Leaves Door Open
A US judge has declined to block Meta from laying off the 26 employees suing it over AI-assisted job cuts. The decision, however, may be revisited once more details emerge about how the technology was used.
District Judge William Orrick, in Oakland, California, rejected the workers’ bid to stop Meta laying them off. He found they could not show that losing their jobs amounted to the “irreparable harm” needed for an emergency order. The plaintiffs, who filed anonymously, comprise engineers, managers, researchers, and designers. Details of the hearing and the ruling were reported by Reuters.
The layoffs, due to begin on 22 July, can proceed while a longer-lasting preliminary injunction remains pending. Orrick indicated at the hearing that he would likely rule on this next month. He also suggested he could change his mind once he has more details about the layoffs.
What the Plaintiffs Say the Ruling Leaves Open
In a joint statement reported by Reuters, lawyers for the workers said that while their request was denied, the judge recognised the case raises “serious questions” about Meta’s conduct. They pointed to the court’s express statement that it may reconsider its determinations “based on any additional evidence the parties provide regarding whether and how AI was used” in the reduction in force.
This suggests the mechanics of Meta’s selection process – what the systems measured, and how the outputs were used – remain under scrutiny, and will be directly at the centre of what comes next.
Meta has denied wrongdoing throughout the proceedings and maintains that decisions involving the layoffs were made by humans.
The Stakes Argued in Court
Counsel for the plaintiffs told the court that alongside jobs and salaries, the workers stood to lose stock options and health insurance, jeopardising medical care for pregnancies and other conditions. “There’s no do-over for bonding with a new baby or giving birth or having active medical treatment,” lawyer Barbara Cowan told Orrick, Reuters reported. Erin Connell, representing Meta, countered that what the workers were losing was employer-subsidised cover.
Timing is pressing. The layoffs are scheduled to be finalised on 22 July for many of the workers, and later in July or August for others.
An Unusual Procedural Route
The case is being fought on two tracks at once. The plaintiffs asked Orrick for emergency relief while pursuing their substantive claims in private arbitration. Their argument is that Meta’s agreements require employees to arbitrate workplace disputes individually, but do not extend to requests for temporary relief.
Such carve-outs are common in arbitration agreements, but they are usually invoked in disputes over alleged trade secret theft or the poaching of clients and staff, not layoffs of at-will employees. Using one to challenge a redundancy process is novel, and part of what makes the case worth tracking.
What Is Being Alleged
The plaintiffs claim Meta used what they describe as “a constellation of internal artificial intelligence systems” to score and rank employees on a termination list. These allegedly include a tool known internally as ‘Metamate’, employee-trained ‘second brain’ agents, keystroke and activity monitoring, AI-token-usage dashboards, and AI-assisted performance ranking and calibration.
They allege Meta did not neutralise gaps created while employees were on holiday or legally protected leave, causing the AI adoption scores fed into layoff selection to fall as a result.
Meta notified nearly 8,000 employees, around 10% of its global workforce, in May that they were losing their jobs, as the company increases its investment in AI.
Why This Case Against Meta Still Matters for HR
Losing an emergency application is not losing a case. The substantive claims are unresolved, and the judge has kept the door open specifically on the question of how AI was used.
Legal experts and HR leaders alike will be watching this case closely. As organisations start to embed AI into decision-making processes, there needs to be clarity on where the line is drawn between the algorithm and human judgement.
Martyn Redstone, Head of Responsible AI at Warden AI, states: “Automated decision systems should never serve as the sole or primary determinative factor in high-stakes decisions like layoffs. An algorithm lacks the capacity for legal compliance, contextual reasoning, or statutory awareness.”
