OpenAI's $3.2M PERM settlement signals DOJ crackdown on hiring bias

For corporate legal departments and employment attorneys, the OpenAI settlement serves as a stark reminder that the DOJ’s Civil Rights Division is closely scrutinizing how companies run their PERM labor certification programs. The $3.2 million penalty and mandatory policy changes highlight the legal risks when recruitment processes are designed to discourage U.S. applicants—even if no discriminatory intent is admitted.

OpenAI and its subsidiary Statsig Inc. have agreed to pay $3.2 million to settle U.S. Department of Justice allegations that they discriminated against American workers in favor of foreign nationals holding temporary visas. The settlement, announced on August 4, 2026, resolves claims under the Immigration and Nationality Act (INA) that the companies structured their Permanent Labor Certification (PERM) hiring processes to discourage qualified U.S. applicants. The DOJ found that OpenAI did not post PERM-related jobs on its own careers website—a stark departure from its standard practice for other openings—required paper applications mailed to a physical address while accepting electronic submissions for other roles, and ran radio advertisements during late-night slots. These practices, according to federal investigators, systematically reduced the pool of domestic candidates, tilting the field toward foreign workers on H-1B or similar visas whom the companies could then sponsor for green cards.

“

The settlement is significant not for its dollar figure—$3.2 million is modest relative to OpenAI’s valuation and revenue—but for the signal it sends about DOJ enforcement priorities in the AI sector.

The settlement is significant not for its dollar figure—$3.2 million is modest relative to OpenAI’s valuation and revenue—but for the signal it sends about DOJ enforcement priorities in the AI sector. Civil Rights Division Assistant Attorney General Harmeet Dhillon underscored the point: “It is illegal to discriminate against U.S. workers by preferring temporary visa holders for jobs.” The DOJ is clearly watching how technology firms, especially those driving critical AI development, balance their voracious appetite for global talent against the legal requirement not to disadvantage U.S. workers. The PERM process is designed to ensure that no qualified American is available before a company can permanently employ a foreign worker, but the DOJ’s findings suggested a deliberate effort to circumvent that safeguard.

OpenAI’s subsidiary Statsig, described as a product software developer, was also implicated, indicating that the problematic practices extended across the corporate structure. The settlement includes $1.2 million in civil penalties payable to the U.S. Treasury and $2 million allocated to compensate affected U.S. workers. Additionally, OpenAI must overhaul its hiring policies, implement mandatory training, and subject itself to DOJ monitoring for a period yet to be disclosed. While OpenAI denied any wrongdoing—a standard posture in such settlements—the compliance terms are intrusive enough to affect day-to-day talent operations.

For the broader technology landscape, this case underscores a persistent tension. AI companies face a limited supply of top-tier research scientists and engineers, many of whom are foreign nationals. PERM labor certification is a crucial pathway for retaining that talent, but aggressive efforts to hire abroad can collide with anti-discrimination laws. The DOJ’s action suggests that even subtle discouragement of U.S. applicants—through archaic application methods or obscure job postings—will be treated as a violation. Other high-growth tech firms, particularly those heavily dependent on H-1B visas, should review their own PERM recruitment protocols immediately. The remedy imposed on OpenAI—publicizing all PERM opportunities widely, accepting electronic applications uniformly, and ending late-night radio ads—may become a de facto compliance template.