US Government Secures 50 Percent OpenAI Discount Under New OneGov Procurement Strategy

The United States General Services Administration has finalized a government-wide agreement with OpenAI to provide public employees with access to the company's advanced artificial intelligence models at a 50 percent discount. Scheduled to take effect on October 1, 2026, the contract establishes a usage-based pricing structure with no minimum spending requirements, expanding the discounted access to federal, state, local, and tribal government entities across the country.

The agreement marks the end of a highly subsidized pilot program that allowed federal agencies to utilize OpenAI's models for just one dollar per year. As the government transitions to a long-term procurement model, the new framework is designed to help public agencies scale their use of artificial intelligence while maintaining strict data security standards.

In August 2025, OpenAI joined the General Services Administration's central purchasing marketplace, known as OneGov, under a pilot program that offered its artificial intelligence models for a nominal fee of one dollar per year. The negotiated terms represented the steepest technology discount ever secured by the federal government's purchasing arm. The pilot program saw rapid adoption across the federal workforce, with approximately 3.5 million federal employees utilizing OpenAI's ChatGPT platform.

According to data released by the General Services Administration, this widespread adoption generated an estimated 1.4 billion dollars in cost savings during its first year. While the new agreement represents a price increase from the initial one-dollar pilot, officials state that the 50 percent discount is structured to allow agencies to transition to a sustainable, consumption-based model. The usage-based pricing system is intended to make it easier for various departments to test, deploy, and scale artificial intelligence tools according to their specific operational needs.

The contract was secured through the General Services Administration's OneGov procurement strategy, a Trump administration initiative designed to leverage the bulk purchasing power of the federal government to secure steep discounts on software and technology acquisitions. By negotiating a single, government-wide framework, the administration aims to eliminate redundant procurement processes across individual agencies.

Data protection and governance serve as central pillars of the new agreement. The contract incorporates the General Services Administration's specific artificial intelligence terms and conditions, restricting model access to secure environments that have already received authorization under the Federal Risk and Authorization Management Program, commonly known as FedRAMP.

General Services Administration Administrator Edward C. Forst emphasized the strategic focus of the procurement framework. "GSA's OneGov strategy is positioning the federal government for the future by integrating advanced, AI-enabled capabilities into agency operations," Forst said in an official statement. "Our agreement with OpenAI provides a secure, governmentwide framework that helps agencies adopt AI at scale and strengthens their ability to carry out mission-critical work." Forst added that the initiative is intended to enhance federal performance, reduce administrative burdens, and allow agencies to focus on delivering direct results for the public.

Under the terms of the agreement, public servants will gain access to OpenAI's most advanced systems, including the company's newly introduced GPT-6 Astra model. Laura Stanton, the Acting Commissioner of the General Services Administration's Federal Acquisition Service, explained that the consumption-based model is a logical progression as agencies move beyond initial experimentation.

"We launched OneGov AI deals because agencies wanted the ability to try new technologies before committing to a lengthy acquisition," Stanton stated. "As agencies increasingly integrate AI into their regular operations, providing consumption-based access is the next logical step." Stanton also extended an invitation to other artificial intelligence developers to engage with the OneGov platform to foster competition and expand choices for public agencies.

OpenAI Chief Executive Officer Sam Altman welcomed the expanded partnership, highlighting the potential for operational improvements across public administration. "Giving public servants secure access to the best AI tools can help government be more efficient, strengthen cybersecurity, and improve the services people rely on," Altman said. He noted that the company is collaborating with the General Services Administration to advance the broader objectives of the administration's national artificial intelligence strategy.

While OpenAI has successfully solidified its position within the federal procurement system, the broader landscape of government artificial intelligence contracting has been marked by intense corporate rivalry and legal disputes. Shortly after OpenAI secured its initial one-dollar deal in 2025, its primary competitor, Anthropic PBC, negotiated an identical one-dollar-per-year agreement. Anthropic aggressively lobbied federal workers, even hosting a prominent promotional event at Washington's Union Station featuring an address by Anthropic Chief Executive Officer Dario Amodei.

However, Anthropic's relationship with the federal government quickly deteriorated due to a dispute with the Department of Defense. The conflict centered on national security contracts, specifically regarding surveillance and autonomous strike guardrails that Anthropic sought to write into its agreements. The disagreement escalated when Defense Secretary Pete Hegseth labeled Anthropic a supply-chain risk, prompting President Donald Trump to issue an order on his Truth Social platform directing all federal agencies to cease using Anthropic's technology.

The presidential directive prompted Anthropic to file lawsuits against the United States government, triggering a complex legal battle over the limits of executive authority in technology procurement. Initially, the General Services Administration issued a directive to remove Anthropic's models from its central artificial intelligence platform. However, the agency was forced to reverse this decision after a federal judge in California issued a preliminary injunction.

Last month, the California judge ruled that the Trump administration must lift its government-wide ban on Anthropic's artificial intelligence tools. The court suggested that while the Pentagon retains the authority to bar specific technologies from its own national security operations, it does not possess the legal authority to unilaterally prevent non-defense agencies from utilizing those same tools. A separate lawsuit remains active in a Washington court, where a three-judge panel has yet to issue a ruling on the administration's legal arguments.

The new usage-based agreement between the General Services Administration and OpenAI is scheduled to run for more than two years following its October 1, 2026 implementation date. The transition will require individual federal, state, and local agencies to establish their own consumption budgets under the pre-negotiated 50 percent discount framework.

OpenAI continues to expand its footprint within national security agencies. In February, the company reached a separate agreement to provide its artificial intelligence tools directly to the Pentagon, operating independently of the broader civilian procurement frameworks managed by the General Services Administration.