OpenAI model price increases for the US government: The $1 annual fee pilot ends and is changed to a half-price discount

Abstract:

According to Bloomberg, a statement from the U.S. General Services Administration (GSA) showed that

OpenAI will terminate a pilot program that allowed government agencies to use its model for only $1 per year and move to a usage-based pricing system

OpenAI

In August last year, in order to promote the popularization of models, OpenAI joined OneGov, the market platform of the US General Services Administration. The platform is a hub that allows federal agencies to purchase AI software at scale. Bloomberg previously reported that the $1 per year clause negotiated in the first year was the largest technology discount negotiated so far by the U.S. government’s central procurement department. The U.S. General Services Administration said that during this period, 3.5 million federal employees used OpenAI’s ChatGPT, saving $1.4 billion in costs.

While the price is an increase from last year’s agreement, the 50% discount is intended to make it easier for government agencies to test and scale AI tools, including the company’s new GPT-6 Astra model. OpenAI will announce on Thursday that it will expand this discount beyond the federal government to include state, local and tribal governments.

This new agreement between OpenAI and the US government will take effect on October 1st and will be valid for more than two years.

Just days after OpenAI reached an agreement with the U.S. General Services Administration last year, its competitor Anthropic also negotiated a deal with the same price of $1 per year. A few weeks later, Anthropic hosted a splashy event at Union Station in Washington to promote its AI tools to federal workers. Anthropic CEO Dario Amodei also delivered a speech in person.

However, at least for Anthropic, U.S. government business was quickly overshadowed by a messy dispute with the Pentagon. The dispute involves mass surveillance and security limits on autonomous strike systems that Anthropic wants to build into contracts focused on national security.

Subsequently, the U.S. Department of Defense classified Anthropic as a supply chain risk. In response, Anthropic sued the U.S. government. Last month, the judge in the California lawsuit ruled that the Trump administration must lift a ban on the use of Anthropic AI by federal agencies. This means that while the Pentagon can ban the use of the technology in its own work, it may not have the power to unilaterally force other government agencies not to use Anthropic's models.