IBM Pays $17 Million Due to Its DEI Practices—Here Are the Accusations

IBM agreed to pay $17,077,043 to the federal government to resolve allegations of illegal diversity, equity and inclusion (DEI) practices.

The DOJ touted the settlement as the first False Claims Act resolution secured under the Civil Rights Fraud Initiative, launched in May 2025. The False Claims Act is a 1863 law that forbids companies from overcharging or defrauding the government. The DOJ argues that engaging in certain DEI practices while holding a federal contract or receiving federal funds can be construed as fraud against the government. IBM is a federal contractor.

{snip}

IBM denies all the allegations, but the misconduct that the Justice Department claims it rooted out, according to the settlement agreement, includes diversifying the pools of candidates for interviews and training and development programs.

{snip}

According to the settlement, the Justice Department alleged that IBM took race and gender into account in its hiring practices by using “diverse slates” and “diverse sourcing.” They argue that IBM used these practices to select who to interview.

{snip}

“Diverse sourcing” was also not specifically defined, but it generally refers to expanding recruiting efforts beyond traditional pipelines to reach a wider range of candidates. In practice, it typically includes outreach to different schools or communities. Again, the goal is to expand opportunity, not dictate hiring outcomes.

{snip}

The settlement also alleges that IBM provided training, development, mentorship, and leadership programs that took race or sex into account. Similar allegations have surfaced at other companies. For example, the EEOC recently filed a federal lawsuit against a Coca-Cola distributor, alleging discrimination against men after the company organized a two-day networking trip for women employees.

The allegations also claim that IBM used a diversity multiplier in its bonus calculations to help meet compensation targets, though it’s unclear what those targets were or how the multiplier was applied. Tracking compensation by race and gender is widely viewed as a useful tool for identifying bias. Gender or racial disparities in pay often emerge from the accumulation of many small decisions, and without aggregate data, those patterns can go unnoticed.

{snip}