Staff at the US House of Representatives cannot utilize Microsoft’s AI-based chatbot, Copilot. This decision was made after the Office of Cybersecurity recognized possible hazards with the application, including the risk of exposing sensitive House data to unapproved cloud providers. The ban removes and blocks Copilot from all House Windows devices.
This decision shows the House’s continuous efforts to regulate the fast developing AI sector and navigate internal AI use. In June of last year, the House confined ChatGPT to the paid membership version under specified conditions and banned the free version.
Microsoft Copilot, developed with ChatGPT developer OpenAI, is a standalone chatbot for online and mobile platforms with commercial versions that integrate directly into Microsoft Office apps including Word, Excel, Outlook, and PowerPoint.
Microsoft plans to release a package of AI capabilities for government usage in reaction to the prohibition to meet federal government security and compliance requirements. These tools should be released later this year.
Catherine Szpindor, the House’s Chief Administrative Officer, stated that the ban extends to commercial Copilot. If a government version of Copilot becomes available, the House will evaluate its suitability for usage. This judgment highlights US authorities’ and companies’ concerns about AI’s security consequences, specifically the risk of confidential data being shared with unauthorized cloud services.