The U.S. House has set a strict ban on congressional staffers' use of Microsoft Copilot, the company's AI-based chatbot, Axios has learned.
Why it matters: It's the latest example of the federal government trying to navigate its internal use of AI while simultaneously attempting to craft regulations for the burgeoning technology.
The House last June restricted staffers' use of ChatGPT, allowing limited use of the paid subscription version while banning the free version.
Driving the news: The House's Chief Administrative Officer Catherine Szpindor, in guidance to congressional offices obtained by Axios, said Microsoft Copilot is "unauthorized for House use."
"The Microsoft Copilot application has been deemed by the Office of Cybersecurity to be a risk to users due to the threat of leaking House data to non-House approved cloud services," it said.
The guidance added that Copilot "will be removed from and blocked on all House Windows devices."
What they're saying: Microsoft hopes the suite of government-oriented tools they plan to roll out this summer will address Congress' concerns.
"We recognize that government users have higher security requirements for data," a Microsoft spokesperson told Axios.
"That's why we announced a roadmap of Microsoft AI tools, like Copilot, that meet federal government security and compliance requirements that we intend to deliver later this year."
What we're hearing: The CAO's office told Axios that the guidance "applies to the commercial version" but that their office "will be evaluating the government version when it becomes available and making a determination at that time."
Zoom in: Copilot is Microsoft's AI-based assistant, built on top of technology from ChatGPT creator OpenAI.
Microsoft has released free and paid consumer versions of the software, as well as a variety of paid options for businesses.
Microsoft's Copilot works as a standalone chatbot for the Web and mobile devices, and the paid versions can also work directly within Office apps like Word, Excel, Outlook and PowerPoint.
Zoom out: The issue facing the House parallels concerns that companies have faced, with many opting to block access to consumer chatbots like ChatGPT for fear of data leakage.
Many businesses are eyeing or purchasing business versions that come with guarantees that data won't be used to train future models — which opens the risk of data leakage.