The General Services Administration has signed a new partnership agreement with OpenAI for the company’s ChatGPT offering, a move the agency says supports President Trump’s AI Action Plan.
ChatGPT Enterprise will be offered to agencies at a steep discount — $1 per agency for the next year.
“GSA is playing a leading role in the Trump administration’s adoption of AI technology by government,” GSA Acting Administrator Michael Rigas said in a release Monday. “Our government’s effective use of AI is critical to demonstrating we are the world’s AI leader, and we are thankful for OpenAI’s partnership.”
This agreement is part of GSA’s OneGov initiative to centralize technology procurement and comes just a day after the announcement that ChatGPT, Google’s Gemini, and Anthropic’s Claude had been added to the Multiple Award Schedule program.
GSA has not responded to a request for comment on whether similar OneGov agreements are in the works for Gemini and Claude.
The push for greater access to AI tools has spread across the government The Defense Department signed a $200 million contract in July with Anthropic, Google, OpenAI and xAI to provide AI-related tools for national security missions.
“One of the best ways to make sure AI works for everyone is to put it in the hands of the people serving our country,” OpenAI CEO Sam Altman said in his company’s OneGov agreement release.
Under the OneGov agreement, OpenAI is making its leading frontier models available. These are the company’s most capable AI systems.
OpenAI is also making educational tools and training available. The company has created a government-user community and tailored introductory training through its OpenAI Academy.
“To help federal agencies make the most of ChatGPT, we’re teaming up with experienced partners Slalom and Boston Consulting Group to support secure, responsible deployment and trainings,” OpenAI said in its announcement.
To address security concerns, OpenAI has said it will not use the data federal employees type into ChatGPT to train future versions of the model. Responses to agency inputs also will not be used to train the model.
In essence, OpenAI has created a data firewall so that agency use remains separate from development of the AI model.
As part of the agreement, GSA has an “authority to use” for ChatGPT Enterprise. ATUs indicate that a product has met security and compliance requirements.
