Meta Denies AI Use in Layoffs Amid Lawsuit Claims

Meta Platforms has denied allegations that it utilized artificial intelligence in its recent layoff decisions, a claim made by 26 former employees in a lawsuit filed last week. The plaintiffs contend that AI tools, including an internal assistant and dashboards tracking AI token usage, were used to select individuals for termination, allegedly impacting those on leave or with disabilities disproportionately. In a court filing on Thursday, Linh Doan, Meta's director of HR Business Partner Enablement, asserted that layoff selections were made by "human business leaders" and that AI played no part in the process. Doan specifically refuted claims that leave status, leave history, disability status, or accommodation requests influenced these decisions, stating that such protected characteristics were not considered.

The lawsuit prompted a judge to request an explanation from Meta regarding the termination of four plaintiffs whose immigration status was at risk, following the denial of a temporary restraining order that would have halted the layoffs. Doan's filing provided insight into Meta's layoff methodology, emphasizing that decisions were driven by human judgment and objective business rationale. She explained that the company first identified organizational areas and specific employee "cohorts" (defined by level and title) slated for layoffs based on business needs. Subsequently, "selection criteria" were developed for evaluating employees within these groups.

Doan further elaborated that these selection criteria were required to be "objective and tied to the business rationale." For instance, if the business rationale focused on retaining top talent within a specific team, the selection criteria would be designed to identify and retain those individuals. This process, as described by Doan, aimed to ensure that layoff decisions were grounded in business objectives and evaluated through a human-led, objective lens, directly countering the plaintiffs' accusations of AI-driven and discriminatory practices. The company's defense centers on the principle that human leaders made the final decisions based on predefined, business-related criteria, not on algorithmic assessments or protected personal information.