OpenAI, Leidos partner to deploy AI for federal agency missions
OpenAI is teaming up with longtime government contractor Leidos to roll out generative and agentic artificial intelligence for specific missions at federal agencies.
The partnership, shared first with FedScoop on Thursday, marks the latest example of AI’s growing presence in federal agencies as the Trump administration encourages use of the technology in government to reduce the time and costs of certain tasks.
The initiative will focus on “more advanced mission areas” of the government, and bring together OpenAI’s models with Leidos’s data, “deployment expertise and their understanding of the mission,” Joseph Larson, vice president of government at OpenAI, told FedScoop in an interview.
As part of this, Leidos will incorporate OpenAI’s ChatGPT and API Platform into its core operational systems, giving every Leidos customer access to the company’s automation, product design, and delivery, according to a company press release. Engineers from both companies will work closely together on mission-specific tasks, Ron Keesing, Leidos’ chief AI officer, told FedScoop in a separate interview.
OpenAI’s ChatGPT product is already available to agencies for $1 through a OneGov deal with the General Services Administration, and the company is in its final stages of the FedRAMP approval process. Larson noted the company’s partnership with Leidos is building upon this, and is geared toward more tailored government tasks.
“What we’re trying to do is move just beyond the provisioning of a ChatGPT Enterprise service into actual, cutting-edge AI for large-scale government missions,” Larson said.
The idea, according to Keesing, originated in part from the federal contracting giant’s internal use of OpenAI tools. Leidos saw an opportunity to expand this work into the public sector, Keesing added, pointing to potential use cases like supply chain monitoring, reporting, analytics, and issue detection.
“We’ll always respond to what our customers ask for, but we are also aware of their mission needs, and we’re actually working together in anticipation of those needs to develop solutions and show them what agentic solutions really make possible today that wasn’t possible before,” Keesing said.
Agentic AI differs from traditional AI in that it can carry out multi-step tasks with limited human oversight, and Keesing said the technology will allow workers to get away from repetitive tasks and focus on “higher value” action items.
As for what prompted Leidos to choose OpenAI as a partner, Keesing credited the familiarity of the products in everyday life.
“This is the kind of work people do that they’ve already discovered and use in their own homes,” he said, pointing to ChatGPT. “And we found that when we introduce it into transforming knowledge work in professional settings, they take to these tools very naturally and they gain a great deal.”
The Trump administration has pushed for agencies to rapidly adopt AI and other emerging technologies to boost federal government efficiency. This has led to a push to reduce regulations, which the administration and several technology companies credit as hindering U.S. competition and progress.
A slew of technology companies have expressed more interest in working with the government under this new perspective.
“You can think about this as us having the opportunity to bring forward the approach we’ve been using for a long time in a more formal way,” Keesing added.
Larson added that OpenAI has seen increased government demand for guidance on applying AI in a “mission context.”
“How do I actually use it to solve the problems that are most critical to me as a government agency?” he said.
This is not the first AI partnership for Leidos. In 2023, the company partnered with Microsoft to increase the use of generative AI in the public sector.