According to Axios, US Congressional staff members are no longer able to use Microsoft’s Copilot on their official devices. According to the magazine, House Chief Administrative Officer Catherine Szpindor sent Congress staff a memo informing them that the AI chatbot is now formally forbidden. It appears that Copilot poses a risk “because of the possibility of House data being leaked to non-House approved cloud services,” according to the Office of Cybersecurity. Though they can still use Copilot on their personal laptops and phones, it will no longer be available on any Windows device that belongs to the Congress.
Almost a year ago, the Congress also set a strict limit on the use of ChatGPT, which is powered by OpenAI’s large language models, just like Copilot. It banned staffers from using the chatbot’s free version on House computers, but it allowed them to continue using the paid (ChatGPT Plus) version for research and evaluation due to its tighter privacy controls. More recently, the White House revealed rules federal agencies have to follow when it comes to generative AI, which would ensure that any tool they use “do not endanger the rights and safety” of Americans.
Microsoft told Axios that it does recognize government users’ need for higher security requirements. Last year, it announced a roadmap of tools and services meant for government use, including an Azure OpenAI service for classified workloads and a new version of Microsoft 365’s Copilot assistant. The company said that all those tools and services will feature higher levels of security that would make it more suitable for handling sensitive data. Szpindor’s office, according to Axios, will evaluate the government version Copilot when it becomes available before deciding if it can be used on House devices.
