A federal judge denied emergency relief to 26 Meta workers challenging allegedly AI-assisted layoff selections, finding no near-term irreparable harm while leaving the discrimination claims, a preliminary-injunction request and the role of Meta's internal tools unresolved.
Meta can proceed with layoffs challenged by 26 workers, but its emergency court victory settles none of the central factual questions about the selection process. The next phase turns on evidence the public accounts do not supply: what each internal measure captured, how protected absences were treated, how those measures affected managers and how the plaintiffs compared with the rest of the layoff pool.
U.S. District Judge William Orrick refused to stop Meta from carrying out the layoffs beginning July 22 because the workers had not shown the irreparable harm required for an emergency order. The underlying claims are due to be decided in private arbitration, while the workers' request for a longer-lasting preliminary injunction remains pending, according to the account of the ruling and hearing.
That is a narrower outcome than a finding that Meta's process was lawful. The workers' lawyers said Orrick could reconsider his determinations if the parties provide more evidence about whether and how AI was used. A separate case preview says a preliminary-injunction hearing is scheduled for August.
The parties disagree about what would be irreparable before arbitration. The workers pointed to lost salary, unvested equity, employer-subsidized health coverage during pregnancy or active treatment, expiring leave rights and immigration consequences. Meta's lawyer argued that employees would lose the insurance subsidy rather than coverage itself and that ordinary employment damages could be recovered if they later prevail. Orrick also requested more information about how Meta selected four employees on company-sponsored visas, according to a limited legal report.
The 26 anonymous plaintiffs include engineers, managers, researchers and designers. Meta said in court filings that affected workers remained on payroll after losing access to company systems on May 20 and had performed no work since then. Final separation was scheduled for July 22 for many and later in July or August for others.
Meta has denied wrongdoing and said workforce and organizational decisions were made by people, not AI. The plaintiffs do not merely allege that a model issued termination orders. They allege that human decisions depended on internal AI systems, activity-monitoring data, AI-token-use dashboards and algorithmically assisted performance rankings that disadvantaged people who could not accumulate the same signals while away from work.
The complaint says Meta failed to account for protected leave or conduct an individualized, leave- and accommodation-neutral review. It names the Metamate large-language-model assistant, employee-trained “second brain” tools and a productivity score allegedly drawn from keystrokes, screen content, emails and browser history among the systems used to score or rank workers. Those descriptions and the causal claim come from the plaintiffs; the public material does not disclose the tools' weights or establish that any one of them determined an individual selection.
That distinction makes Meta's human-decision defense incomplete but also leaves the workers with a substantial proof burden. A manager can rely on a distorted measure, yet the presence of an algorithm does not prove distortion, disparate impact or causation. The evidence would have to connect leave or disability to a measured input, the input to a comparison or ranking, and that output to a termination decision.
All 26 plaintiffs took protected leave or requested or received a disability accommodation, the independent account of the complaint says. Thirteen had caregiving- or pregnancy-related absences: eight women took maternity or pregnancy leave, four men took parental leave, and one woman took family-care leave and later bereavement leave.
Those figures show why this group brought the case; they do not measure the selection system's effect across Meta. The public record does not provide the number or share of all employees in the relevant comparison groups who took protected leave, received accommodations or were selected for layoffs. Without those denominators, the plaintiffs' composition cannot by itself establish that leave-takers, disabled workers or women were selected at a higher rate.
The scope gap is large. Meta said nearly 8,000 employees, about 10% of its global workforce, were notified in May that they would lose their jobs. Its first-quarter regulatory filing reported 77,986 employees as of March 31, up 1% from a year earlier. The 26 plaintiffs are therefore a small, self-selected subset of the planned cuts, not a statistical sample of them.
The complaint alleges violations of federal and state protections, including the Family and Medical Leave Act, the Americans with Disabilities Act, the Pregnancy Discrimination Act and the Pregnant Workers Fairness Act. Federal technical assistance on employment selection explains that a facially neutral procedure can violate anti-discrimination law when it disproportionately excludes a protected group and cannot be justified under the governing standard. It also says disability-screening tests must be job-related and consistent with business necessity, and should measure the intended skill or factor rather than an impairment. Applying those principles here still requires evidence about Meta's actual procedure and the affected populations.
The case appears to be the first against a major U.S. company centered on alleged AI use in choosing employees for layoffs. It is not the first effort to apply established employment law to algorithmic workplace screening.
Job applicants have accused Workday of discrimination based on race, age and disability through algorithm-based screening tools. A March 2026 Northern District of California order records that disparate-impact claims had survived two rounds of dismissal motions, then illustrates how claim-specific the litigation remains: the judge declined to dismiss the applicants' federal age disparate-impact theory, but dismissed one plaintiff's physical-disability theory and the California claims at issue, with permission to amend.
That procedural history is not a finding that Workday discriminated, and applicant screening is not a reduction in force. Its relevance is limited but useful: courts can evaluate algorithmic selection under older discrimination rules, while each protected group, challenged practice, causal link and jurisdiction still needs its own proof.
The preliminary-injunction proceeding is the next public decision point, although some separations may be complete before the hearing. The court can revisit emergency findings on a fuller record; the ultimate employment claims are set for individual arbitration under Meta's agreements.
Four categories of evidence would sharpen the central question:
A leave-neutral comparison could test whether absence itself depressed the plaintiffs' standing. Managerial records could show whether human review corrected, preserved or ignored any effect. Until that evidence emerges, the order authorizes Meta to proceed with the layoffs; it neither validates the selection process nor proves the workers' theory.
Christian Faith Madison’s estate alleges GPT-4o sustained a months-long narrative of prophecy, sacrifice and resurrection before her death. Public safety disclosures and controlled research make that failure mode plausible, but neither establishes which model behavior she encountered or whether it caused her death.
Shanghai AI Laboratory has released Intern-S2-Preview-397B as downloadable weights and a hosted API, but its own 35B alternative offers the same context window and a far smaller disclosed scale while the larger model still lacks auditable comparative scores, pricing and physical-validation results.
Mixfont's free Decoy Font gives each glyph a sharp decoy and a blurred intended letter, creating a hurdle for pixel-based readers while leaving selectable text, known-image techniques and accessibility costs outside its protection.
Zhipu reportedly reached $1 billion in annual recurring revenue in July, roughly four times a March estimate, but the unconfirmed run rate is not annual sales and still sits far ahead of recognized cloud revenue while margins remain thin.
Cmpunlocker says patched Nvidia open kernel modules can expose 64GB of HBM on an 8GB CMP 170HX or 40GB on a 10GB card while removing an SM throttle. The exploit mechanism is documented, but the retained evidence does not independently validate the tool across full-memory workloads, error rates, power use or A100-class performance.
An unidentified industry source says at least one Nvidia board partner has physical RTX 50 Super cards but cannot sell them, with 3GB GDDR7 quoted at $60 to $70 per chip. The reported memory bill explains a plausible pricing problem, but neither Nvidia nor a second independent source has confirmed the products, the hold or the component terms.
Anthropic’s 20-year lease gives TeraWulf a customer for 401 megawatts of future AI infrastructure, but rent starts only after delivery and the proposed utility deal passes power and grid costs to TeraWulf, leaving financing and project margin unresolved.
OpenAI has put conversations and Projects back in its redesigned ChatGPT desktop app and enabled cloud Work threads to move across devices, correcting the launch's biggest usability failures without merging local Work or Codex histories.
Twenty-nine countries signed an agreement creating WAICO as an independent intergovernmental organization, while China paired the launch with capacity-building offers that are not yet confirmed as WAICO programs.
Meta reportedly plans to put departing AWS compute executive Dave Brown to work on its data-center buildout, adding hyperscale operating experience while leaving any customer-facing cloud business conditional and undefined.
Moonshot AI has made Kimi K3 available through its apps and API, pairing a 2.8-trillion-parameter architecture with early frontier-level results, but the model's open-weight claim cannot be tested until its weights and technical report arrive.
CIA Director John Ratcliffe said US intelligence is consistent with an estimate that Russian recruits last 20 to 30 minutes on Ukraine’s battlefield, but the public trail leads to an unsourced claim about assault troops and does not establish a representative average.
China has paired a five-year AI training offer for developing countries with cooperation centers, a weather-warning rollout and a new 29-country organization. The package gives Beijing a platform for influence, but no budget, selection rules or delivery timetable has been published.
The FastFlowLM team has joined AMD after building a Ryzen AI NPU inference flow on technology AMD already developed and distributed. The move could reduce model-support delays, but its financial significance, transaction perimeter and performance benefit remain undisclosed.
Federal intervention changed the launches of Anthropic's Claude Fable 5, Claude Mythos 5 and OpenAI's GPT-5.6, but through different legal and voluntary channels. The result is real government leverage over early customers without a common security threshold or durable review process.
A proposal developed with Treasury Secretary Scott Bessent would put an independent AI regulator under SEC oversight, but President Trump had not reviewed it and its tests, funding and enforcement authority remained unresolved.
San Francisco has demanded that Apple and Google cut off 13 apps capable of producing nonconsensual sexual deepfakes. The threatened case could test California’s new liability rules, but the confidential app list and the law’s “primary purpose” definition leave a central question unresolved.
Anthropic will keep Claude Fable 5 inside Max and Team Premium plans from July 20 at 50% of their usage limits, while Pro and Team Standard customers move to separately billed credits. The durable change is who gets bundled access, not a larger allowance for top-tier users.
HKT plans a 3.2 Tbps data-centre route from Lok Ma Chau to Tseung Kwan O by the end of 2026, but has not disclosed customer bandwidth, price, project cost or end-to-end latency, while Sandy Ridge's developer expects to start within 42 months of its land award.
Apple briefly topped Nvidia during Friday trading before Nvidia closed about $6 billion ahead. The filings show why that ranking is a weak AI verdict: Apple’s operating momentum predates its new Siri, while Nvidia’s rapid growth comes with concentrated customers and large forward commitments.