AI Layoff Decisions Face Their First Major Workplace Test

Last updated: November 4, 2026

A lawsuit filed by 26 former Meta employees alleges that AI-enabled systems helped determine who was laid off in a major round of cuts. The case, reported by Fast Company, puts a blunt question into the hiring cycle: can workplace data influence job termination without employees knowing how it is being used?

For professionals, the dispute is a warning about the growing gap between what employers collect and what workers can inspect. Performance dashboards, absence records and AI-generated reviews may appear administrative. In practice, they can become evidence that shapes careers.

• The lawsuit was brought in July by 26 anonymous plaintiffs.
• It alleges that AI-enabled systems were used to select employees for layoffs.
• The plaintiffs say disabled employees needing accommodations and parents took disproportionate hits.
• Meta says humans were involved and that AI did not select anyone for termination.
• More than two-thirds of poll respondents feared employers might use AI to choose layoff targets.

What the Meta lawsuit alleges

One former employee spent five years receiving strong performance ratings, including a rare equity award. Her rating never fell below “exceeds expectations.” After taking medical leave in late 2025, she returned to find that her rating had changed.

Her manager later said the change was needed to fit a bell curve. Her responsibilities also shifted. Meanwhile, dashboards that had ranked her highly for AI adoption had not adjusted for the fact that she was on protected leave.

A former Meta director made a similar point in legal filings. He had monitored his team’s AI usage closely and said the dashboards failed to distinguish leave from low engagement. If the systems were still recording participation while an employee was absent, the resulting rankings could become misleading inputs in employment decisions.

The 26 plaintiffs argue that the effects were broader. They contend that the layoffs disproportionately affected people who needed disability accommodations or had taken parental leave.

Meta’s defence and the wider employment system

Meta has denied the lawsuit’s central allegation. HR leader Linh Doan said AI did not select anyone for termination or assist with selection decisions. The company says people made those choices.

The case nevertheless lands in a workplace already using automated tools for employment decisions. Fast Company reported that major companies including JPMorgan Chase, Disney and Meta use AI to draft self-assessments and performance reviews. People analytics systems also commonly track absenteeism and sick days.

That does not prove AI chose any particular layoff target. It does expose a weak transparency layer. Workers may see performance scores and rankings without knowing how absences, role changes or management adjustments enter the calculation.

Several US states have passed legislation requiring greater transparency about AI’s role in termination decisions, according to Fast Company. Litigation involving HR technology companies such as Workday and Eightfold also shows that automated employment tools are already being challenged in other parts of the employment cycle.

Our analysis: look at the data chain

The important issue is not whether a model pulled a final button. Employment decisions are usually a chain: dashboards produce signals, managers interpret them, HR systems combine records and senior leaders approve the final list.

Each step can introduce distortion. A leave period may depress engagement data. A bell curve may override earlier ratings. A changed job description may make old output look weak. AI can make these inputs faster and more polished without making them fair.

For job advice, the lesson is straightforward. “Performance management” is not a black box you should accept without inspection. Professionals should know what their employer measures, whether leave is excluded, how ratings are adjusted and who can challenge them.

There is a legitimate business case for using data to identify underperformance or allocate work. The problem arises when the data is treated as a clean verdict. A lower score may describe a system rather than a person.

A practical example for an employee

Say you return from parental leave and notice that your collaboration score has fallen. Before accepting that as an objective judgment, ask for the underlying measures. Which activity was counted as engagement? Were you evaluated against the same role you held before leave? Was the score adjusted for time away?

Save the relevant emails, rating summaries and job descriptions. If the explanation is vague, ask for the review process in writing. This is not automatically a dispute. It is basic career-risk management.

What to watch next

Watch whether the case produces clearer evidence about Meta’s selection process and whether state transparency laws force employers to disclose more. The plaintiffs’ allegations remain claims, not established facts.

For workers, the practical stance is neither panic nor complacency. Treat automated performance data with the same attention you would give any record that can affect your career. Ask what it measures before it starts measuring you.

Key takeaways

• Performance data can distort rankings when leave and role changes are not handled properly.
• Ask how AI influences reviews, rankings and termination workflows.
• Keep written records when scores change after returning from leave.
• Treat transparency laws and related lawsuits as signs that workplace AI governance is still developing.