“The opportunity to incorporate generative AI into government work is akin to giving a personal computer to every worker,” says Stephen Ehikian, GSA’s Acting Administrator and Deputy Administrator. “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.”
Lambert here: No, it’s really not.
UPDATE: The General Services Administration (GSA) today confirmed the internal launch of an AI chatbot “designed to support staff in their daily work,” it says.
The unnamed chat tool and API are still a work in progress. “GSA is actively seeking feedback from staff to refine its capabilities and…working towards offering the tool as a shared service to other federal agencies in the near future,” according to the agency.
It opted for an in-house AI rather than a commercial solution from the likes of OpenAI or Google since those solutions “may not meet the security and privacy requirements necessary for government use,” the GSA says.
Original Story (3/10):
Elon Musk’s Department of Government Efficiency (DOGE) has deployed a custom AI chatbot at the General Services Administration, Wired reports, citing sources familiar with the matter.
DOGE began testing the GSAi chatbot with 150 GSA employees in February and it has now rolled out to 1,500 agency employees. It was under development for several months before DOGE arrived, but DOGE accelerated the rollout, sources tell Wired.
GSAi’s default AI model is Claude Haiku 3.5, but users can switch to Claude Sonnet 3.5 v2 or Meta Llama 3.2. Employees were told in an internal memo that they could use GSAi to draft emails, create talking points, summarize text, and write code. The tool was designed to be safe for government use, though employees have been warned against entering sensitive data.
Lambert here: I wonder how it worked out.
Add new comment