The General Services Administration is launching a new artificial intelligence tool for government use, designed to support staff at the agency in their regular, daily work, with the goal of rolling it out to other federal agencies in the near future.
GSA began building the technology 18 months ago under the Biden administration, focusing on the security and privacy of the AI tool, a GSA spokesperson tells NBC News.
The decision to develop the tool internally was driven by concerns within the department about the risks of using commercially available AI tools, according to a GSA official familiar with the development of the tool. The tool was ultimately built internally to ensure it meets government security requirements.
GSA is working to see how the AI tool fits into the daily workflow of federal works and is encouraging — but not requiring — staff to use the tool, according to the official.
"The opportunity to incorporate generative AI into Government work is akin to giving a personal computer to every worker,” GSA’s Acting Administrator and Deputy Administrator Stephen Ehikian touted the new tool, in a release, saying, “We are just at the start of our journey using this new tool, but the demand for this technology exists across GSA and the broader government.”
The Trump administration has been supportive of the self-developed AI tool, the spokesperson tells NBC, noting that while the tool is still in its early stages, the rollout within GSA will allow the government to get the feedback necessary to make adjustments that will improve the functionality and how quickly it can be delivered to other agencies.
Tech mogul Elon Musk, a top adviser to President Donald Trump, has pledged to make the government more efficient in part by embracing AI.
Musk's Department of Government Efficiency had been expected to use an AI tool to evaluate answers to an email sent to federal workers asking them to justify their positions, three sources with knowledge of the system told NBC News. Musk at the time denied in a post on X that an AI tool was evaluating the answers. "This was basically a check to see if the employee had a pulse and was capable of replying to an email," he said.
