Learn more. This feature uses an automated voice, which may result in occasional errors in pronunciation, tone, or sentiment. (Getty Images) Employees at the General Services Administration appear poised to test Grok 3, the artificial intelligence tool built by Elon Musk’s company xAI, according to a conversation on a GitHub page referencing the agency’s work. A GitHub page, called “10x-ai-sandbox” and apparently operated by GSA and its digital government group Technology Transformation Services, now references the Grok AI model. One section is called “Introduce Grok 3” and appears to be somewhat active, showing several GSA staffers discussing the tool. “GSA is evaluating the use of several top-tier AI solutions to empower agencies and our public servants to best achieve their goals,” a GSA spokesperson told FedScoop in a response to an inquiry about the agency’s work with Grok. “We welcome all American companies and models who abide by our terms and conditions.” A post from Tuesday shows what appears to be one GSA employee trying to access Grok 3 for testing, but struggling to do so. Several names of the people active on the GitHub page match those of workers affiliated with GSA. Advertisement Another section references launching several models, including versions of Anthropic’s Claude and Meta’s Llama tools, in addition to discussion about the introduction of Grok 3. The 10x AI Sandbox project is described on GitHub as “a venture studio in collaboration with the General Services Administration (GSA). Its primary goal is to enable federal agencies to experiment with artificial intelligence (AI) in a secure, FedRAMP-compliant environment.” It continues: “By providing access to base models from leading AI companies and offering advanced UI features, the sandbox empowers agencies to test and validate new AI use cases efficiently.” Advertisement Interest in testing Grok comes as GSA continues to work on GSAi, an artificial intelligence tool built by the agency and meant to help employees access multiple AI models. At launch, the GSAi tool included access to several systems, including tools from Anthropic and Meta. Notably, Grok came under fire this week after promoting various antisemitic statements on the Musk-owned social media platform X. Two people familiar with the matter told FedScoop that Grok had recently been approved to be added to the models available through GSAi. Another official not familiar with the inclusion of Grok emphasized that the GSAi is designed to provide users with access to a range of models and isn’t designed to make someone use any one particular technology — including a new model is akin to adding an option to a menu. A tool being in the system doesn’t, on its own, mean it’s a great model to use, they said. Work on the GSAi tool has also been discussed within the Chief Artificial Intelligence Officers Council, which convenes AI leaders across government regularly. Last month’s meeting included discussion of GSAi, two people said. An additional person familiar with a meeting that had CIOs and CAIOs in attendance said that GSAi was an agenda item. The General Services Administration is currently exploring potential integration with GSAi and technology being built by or already in place at other federal agencies, two sources told FedScoop. That effort, the people said, is related to AI.gov, a not-yet-launched governmentwide AI platform. Advertisement In February, Wired reported that DOGE was working on a government chatbot called GSAi through the GSA. The following month, GSA announced that it had developed a new chatbot meant for federal workers. At the time, GSA told FedScoop the tool had been in development for several months and involved workers from both Technology Transformation Services — led by DOGE-affiliate and former Tesla engineer Thomas Shedd — and IT workers at GSA. 404 Media later reported that GSA seemed to be involved in building a governmentwide AI tool and that the government had repurposed AI.gov — a domain previously used for AI use case inventorying — for the chatbot. In a possibly related action, GSA also recently registered the domain usai.gov. AI.gov was supposed to go live by July 4, according to 404 Media, but the website still redirects to the main White House page. Online chatter on GitHub shows that back in April, GSA was hoping for formal White House approval to use the AI.gov brand identifier. Written by Rebecca Heilweil and Madison Alder In This Story