"U.S. Congress Implements AI Restrictions: Copilot Banned on Government Devices"
The U.S. Congress has taken a significant step in tightening the use of artificial intelligence technologies within its operations. Staff members are now prohibited from using Microsoft's AI chat-bot, Copilot, on government-issued devices. This development, reported by Axios citing a letter from the House of Representatives' Chief Administrative Officer, Catherine Szpindor, underscores growing concerns over cybersecurity and data privacy within the legislative body.
The decision to ban Copilot appears to be driven by the Cybersecurity Office's assessment that the AI tool poses data leak risks, potentially compromising the House's information to cloud services. While the use of Copilot on personal devices such as smartphones and laptops by Congressional employees remains unrestricted, the move signals a cautious approach towards AI technologies that interact with sensitive information.
Microsoft, in response to Axios, acknowledged the necessity for heightened security measures for government users. In 2023, the tech giant announced plans to develop tools and services tailored for government use. This includes the Azure OpenAI service for classified workloads and a new iteration of the Copilot for Microsoft 365, both promising enhanced security features suitable for handling confidential data. Szpindor's office is expected to evaluate the government version of Copilot once available, before making a decision on its use on Congressional devices.
This policy follows a similar restriction imposed nearly a year ago on the use of ChatGPT, another AI based on OpenAI's large language models, akin to Copilot. However, Congress allowed the use of ChatGPT Plus, a paid version of the AI with stricter privacy controls, on its computers. This series of measures reflects a growing trend among governmental bodies to balance the benefits of AI technologies with the imperative of safeguarding sensitive information against potential cyber threats.
