Protected Leave & AI Selection.​

​​

Why timing matters.

​​

Where an employer uses neutral performance metrics to select employees for termination, those metrics are not actually "neutral" when the evaluation window overlaps with an employee's protected leave. An employee who was on medical leave, parental leave, military leave, or other protected absence during a calibration window cannot demonstrate — by definition — the same deliverables, roadmap ownership, manager sponsorship, or "future-skills readiness" as continuously-present peers. Courts have recognized for decades that when protected leave prevents an employee from meeting facially neutral criteria, the criteria must be adjusted. An algorithm that ingests the same metrics without correction can produce the same problem at scale and speed.

​

The same analysis applies across categories of protected leave. FMLA-qualifying leave, pregnancy and pregnancy-related disability leave, leave as a disability accommodation under the ADA, USERRA-protected military leave, state-law paid family leave, and paid sick leave each carry protections against retaliation and interference. Where a selection process disadvantages employees who exercised any of these rights, substantial legal concerns arise.

Why reaching out early can matter

​

Once large-scale terminations are finalized, it becomes significantly harder — practically and doctrinally — to restore employees to their roles, to obtain information about how selection decisions were made, and to preserve the evidence necessary to challenge the process. Employees who come forward before an adverse action is finalized may be positioned to seek remedies — including, where the legal standard is met, emergency injunctive relief — that are not practically available afterward.

​

If you are currently employed at Meta and have taken protected leave during the past 12–24 months, a consultation now can preserve options that may not be available later.

​

Performance-review signals to watch for.​

​

In matters involving AI-assisted or calibration-driven performance review processes, the following patterns are among those that can indicate a selection process has not adequately accounted for protected leave:

​

• Calibration comparisons made to full-year peers without adjustment for leave-shortened evaluation periods.

• Performance expectations held constant despite the duration of protected leave.

• Ratings trajectories that changed during or shortly after protected leave.

• Retention outcomes tied to recent deliverables, current roadmap ownership, or manager sponsorship accrued during a window in which the employee was on leave.

​

Structural signals to watch for.

​

Beyond the review mechanism itself, structural patterns in a workforce reduction can also raise concerns about leave-status awareness in selection:

​

• Disparity signals at the team level (e.g., a pattern of employees who were on leave appearing on the list).

• Reassignments, reclassified responsibilities, or changed reporting lines implemented while an employee was on leave.

• Permanent redistribution of an employee's scope and hiring of replacements during leave.

• Decisions about an employee's future made during their leave rather than after return.

​

Keystroke tracking and AI training.

 

In parallel with its investigation of AI-assisted selection, Lumen Law Center is investigating employee accounts indicating that Meta is collecting detailed keystroke, screen, and activity data from employees — including director-level and managerial personnel — and using that data as an input to artificial intelligence systems intended to assess, reduce, or replace the workforce generating the data.

​

Workplace monitoring is not new. What is new is the convergence of three practices at scale: pervasive activity logging applied uniformly across exempt and non-exempt roles; undisclosed or under-disclosed secondary use of that data to train AI models; and the deployment of those models in decisions about staffing, performance, and workforce reduction. Where an employer captures the detailed work product of its employees and channels that work product into systems designed to assess, reduce, or replace those employees, serious questions arise under federal and state privacy, labor, employment, and consumer-protection law — and under the contractual and policy commitments the employer has made.

​

Note: Descriptions on this page reflect the firm's ongoing investigation as of April 2026 and are not findings of fact. Specific facts of any individual's claim would be assessed during a confidential consultation.

​

Potential claims: the legal framework

 

Depending on jurisdiction, job classification, and the specific practices at issue, the conduct under investigation may implicate a range of statutory and common-law protections. The categories below are illustrative, not exhaustive, and any individual employee's situation requires specific analysis.

​

I.

FMLA Interference & Retaliation

Interference under 29 U.S.C. § 2615(a)(1) where employer practices discourage or penalize the exercise of FMLA rights; retaliation under § 2615(a)(2) where adverse action follows protected leave; claims where facially neutral selection criteria penalize employees unable to accrue measured "contributions" during a protected-leave period.

​

II.

Protected Leave & Accommodation

Claims under the ADA, the Pregnant Workers Fairness Act, USERRA, state disability-leave statutes (including California's CFRA and PDL), and paid-family-leave laws where AI-driven selection fails to adjust for, or affirmatively disadvantages, employees who took protected absence.

​

III.

Discrimination & Disparate Impact

Disparate-impact claims under Title VII, the ADEA, the ADA, and state civil-rights laws where AI systems trained on productivity data produce discriminatory outcomes; opacity issues under emerging AI-in-employment laws (NYC Local Law 144, California and Illinois AI-hiring regulations, and similar measures).

​

IV.

WARN Act & State Analogs

Compliance with the federal WARN Act and state analogs (including California's Cal-WARN) where mass layoffs or plant closings trigger notice obligations; severance and benefit-continuation considerations that arise when WARN-covered events proceed without adequate notice.

​

V.

Privacy & Data Protection

Intrusion upon seclusion where monitoring extends to personal accounts or devices; collection, retention, and secondary use of personal data without adequate notice or legitimate business purpose; state consumer-privacy violations (including CCPA/CPRA where applicable) and breach of purpose-limitation principles when data collected for one reason is repurposed to train AI.

​

VI.

Electronic Communications & Monitoring Notice

Issues under the Electronic Communications Privacy Act and state analogs (including California Penal Code § 631); violations of state electronic-monitoring notice laws (New York Civil Rights Law § 52-c, Connecticut Gen. Stat. § 31-48d, Delaware 19 Del. C. § 705) requiring clear prior notice of the nature, scope, and purposes of monitoring.

​

VII.

Labor & Concerted Activity

Potential interference with Section 7 rights under the National Labor Relations Act where surveillance chills protected concerted activity; retaliation based on data-derived inferences about organizing, complaint-making, or whistleblowing; use of monitoring to surface and punish protected communications.

​

VIII.

Confidentiality & Privilege

Capture of attorney-client privileged communications, board-level strategic deliberations, and third-party confidential information through keystroke logs and screen recordings; exposure of that material to AI-training datasets; fiduciary and professional-responsibility concerns for officers, directors, and licensed professionals.

​

IX.

Contract, Handbook & Policy Violations

Breach of employment agreements, executive agreements, equity documents, and privacy notices that promised limited monitoring or restricted uses of employee data; unfair or deceptive practices where actual monitoring or AI-training uses diverge from stated policies; breach of the covenant of good faith and fair dealing.

​

X.

Wrongful Termination & Adverse Action

Termination, demotion, or performance-action claims where adverse decisions were driven by opaque AI-generated assessments; pretext theories where "productivity" metrics mask discriminatory or retaliatory motives; claims arising from use of surveillance or selection data in severance leverage or release negotiations.

​

XI.

Lost Equity & RSU Forfeiture

Where a termination is unlawful, the resulting forfeiture of unvested equity is not a technicality — it is part of the injury. Lost RSUs, stock options, and refresh grants may be recoverable as compensatory damages, particularly where termination timing aligns suspiciously with vesting cliffs, refresh events, or bonus dates. At-will employment and arbitration agreements do not eliminate these claims, where a termination has been wrongful. (Please note: without wrongful termination, lost equity and RSU forfeiture are not standalone claims.) 

​

Steps to take now — especially if you are still employed.

 

The actions you take in the days immediately preceding or following an adverse employment decision can materially affect your ability to bring a claim — and, where the legal standard is met, to seek emergency relief. The guidance below is general; specific advice requires an attorney familiar with your situation.

​

• Reach out early if you have concerns.  If you have reason to believe you may be affected by an adverse employment action — particularly if you took protected leave during a relevant evaluation window — a consultation before termination or final action can preserve options that are not available afterward.

​

• Preserve the paper trail.  Save copies of monitoring notices, consent forms, handbook updates, privacy notices, executive agreements, offer letters, leave correspondence, return-to-work communications, performance reviews, calibration feedback, and any internal communications referencing AI, analytics, productivity scoring, or workforce planning. Preserve copies through means you personally control.

​

• Document your leave period and its handling.  If you took protected leave, record the dates, the accommodations or return-to-work arrangements discussed, changes to your scope or reporting line during leave, and any communications about your performance during or immediately after leave. These details matter.

​

• Document adverse actions on a timeline.  Keep a contemporaneous record of performance reviews, changes in duties, PIP placements, metric-based critiques, and any references to system-generated data in employment decisions. Dates and exact language matter.

​

• Protect privileged and personal communications.  To the extent possible and consistent with company policy, do not use monitored systems for privileged consultations or personal matters going forward. Do not, however, make unilateral changes to monitoring tools or your devices before consulting counsel. Do not speak to any AI tool regarding your legal questions, as this can risk attorney-client privilege waiver.

​

• Do not sign severance without review.  If you have been offered a severance agreement in connection with a reduction or performance action, do not sign before obtaining legal review. Severance agreements frequently contain releases, non-disparagement, cooperation, and reference clauses that can foreclose claims you may not yet know you have.

​

• Contact counsel before acting on suspicion.  Do not access restricted systems, disable monitoring tools, take documents you are not authorized to take, or post publicly about your concerns before obtaining legal advice. Each of these can create exposure that a lawyer can help you avoid.

​

​

Confidential consultation.

 

If you are a current or former Meta employee with concerns about AI-assisted layoff selection, treatment during or after protected leave, workplace monitoring, or adverse action based on system-generated data, we want to hear from you. Do not include sensitive personal or employer information in your initial contact — we will provide secure channels for detailed discussion.

​

To request a consultation, email info@lumenlawcenter.com or call / text 310-269-6739.

​

Advertising material. This page is intended for general informational purposes and to describe the services of Lumen Law Center, P.C. It does not constitute legal advice and does not create an attorney-client relationship. Legal rights vary by jurisdiction and depend on specific facts, including the nature of the practices at issue, your job classification, your jurisdiction, the timing of any protected leave, and the terms of any agreements with your employer. Prior results do not guarantee a similar outcome. For advice specific to your situation, contact Lumen Law Center, P.C. directly.