Advertisement

FAA has an agreement for OpenAI technology

The aviation agency has an $80,000 award for “Azure OpenAI CDO,” per USASpending.
Listen to this article
0:00
Learn more. This feature uses an automated voice, which may result in occasional errors in pronunciation, tone, or sentiment.
The Federal Aviation Administration (FAA) building is seen at 600 Independence Avenue in Washington, D.C. (Photo by Eric BARADAT / AFP).


The Federal Aviation Administration appears to have purchased OpenAI technology, according to a delivery order listed on a public repository of U.S. government spending records.

The work, which started last March, is scheduled to end this August, but other details aren’t clear. The agreement simply states “AZURE OPENAI CDO,” likely referencing the OpenAI technology available through Microsoft’s Azure cloud program, which is how the AI company has typically provided its technology to federal customers. The current award amount is just over $80,000, according to the USASpending site

According to a Microsoft website, Azure OpenAI provides access to several of the AI companies’ models, including GPT 3.5-Turbo, GPT 4.0, as well as tools like Whisper and Dall-E. 

OpenAI’s partnership with Microsoft has been critical in helping the company bring its large language models to government users: Through Azure cloud, GPT-4o is now cleared for top-secret use, which means the intelligence community can access the technology. Last August, Azure OpenAI received FedRAMP High authorization. 

Advertisement

The FAA hasn’t responded to FedScoop after nearly a month of inquiries about the business.

OpenAI and Microsoft did not provide a comment, nor did General Dynamics Information Technology, which is named on the USASpending listing. It’s not clear why GDIT is involved, but external companies often help the government with integrating technology packages. NASA, the National Gallery of Art, and USAID, have all accessed OpenAI technology through other firms. 

The FAA has expressed cautious interest in generative AI. An October 2024 research plan released by the agency describes beginning work to “research the use of Large Language Models (LLMs) (such as ChatGPT) in the development and verification of safety-critical software.” 

That commitment seems to show a new attitude toward the technology. In July 2023, the FAA said that it did not use ChatGPT in any of its systems after previously disclosing in its AI use case inventory that its air traffic organization was interested in using the technology for “code writing assistance.” After FedScoop asked why the technology was listed in the agency’s AI disclosure, the FAA removed it and said the entry was made in error. 

The aviation agency has also said that it has used an LLM to answer questions about aviation incident safety reports. MITRE, the federally funded public interest research nonprofit, had also briefly looked at ChatGPT technology while conducting research on transcription tools for the FAA, FedScoop reported last year. 

Advertisement

It’s possible that the FAA could incorporate more generative AI as it pursues IT modernization efforts and amid ongoing concerns about the state of air traffic technology.

OpenAI remains interested in selling to the federal government. The agency recently released ChatGPT Gov, which allows federal agencies to self-host the technology. OpenAI is also pursuing FedRAMP authorization of its own, which would allow the company to work with government customers with the help of Microsoft. 

This story was supported by the Tarbell Center for AI Journalism.

Rebecca Heilweil

Written by Rebecca Heilweil

Rebecca Heilweil is an investigative reporter for FedScoop. She writes about the intersection of government, tech policy, and emerging technologies. Previously she was a reporter at Vox's tech site, Recode. She’s also written for Slate, Wired, the Wall Street Journal, and other publications. You can reach her at rebecca.heilweil@fedscoop.com. Message her if you’d like to chat on Signal.

Latest Podcasts