Key Takeaways
- The U.S. Senate has approved the use of AI chatbots ChatGPT, Google Gemini, and Microsoft Copilot for official tasks.
- Staff can utilize these tools for drafting documents, editing, summarizing information, and preparing briefing materials.
- Microsoft Copilot integrates with Microsoft 365, ensuring data security within the Senate’s environment.
- Not all AI chatbots are approved; Anthropic’s Claude is still under evaluation.
- The Senate’s policy differs from the House of Representatives, which includes Claude among the authorized tools.
AI chatbots approved for U.S. Senate use now include several widely known generative AI tools. The U.S. Senate has authorized staff to use ChatGPT, Google Gemini, and Microsoft Copilot for official work. The approval was communicated through a memo from the chief information officer of the Senate sergeant-at-arms. The policy provides a framework for how Senate offices may use artificial intelligence tools during daily operations.
Senate Allows ChatGPT, Gemini, and Copilot
The approved tools include OpenAI’s ChatGPT, Google’s Gemini, and Microsoft’s Copilot. These chatbots can assist Senate staff with several office tasks. Staff members may use them for drafting documents, editing written material, summarizing information, and preparing briefing materials. The tools may also support research and analysis related to policy work.
The authorization establishes a formal policy for using AI in Senate offices. Reports indicate that some staff had already experimented with these tools before the official approval. The new policy provides guidance and boundaries for how the technology should be used.
Microsoft Copilot Integrated With Senate Systems
Microsoft Copilot has specific advantages for Senate offices. The chatbot operates within Microsoft 365, the software environment used by Senate staff. This integration allows Copilot to function inside existing office applications.
The memo stated that information processed through Copilot remains inside the Microsoft 365 Government environment. This system applies the same security protections used for other Senate data. The arrangement allows staff to access AI features while maintaining government security standards.
Other AI Chatbots Still Under Evaluation
Not all generative AI tools received approval under the Senate policy. Anthropic’s Claude chatbot was not included in the list of authorized systems. Senate information technology officials described the platform as still undergoing evaluation.
The policy in the Senate differs from the U.S. House of Representatives. The House allows four AI tools for official use: ChatGPT, Gemini, Copilot, and Claude. The Senate currently permits only three of those systems.
Government officials continue to study how artificial intelligence should be used in legislative offices. One major concern involves preventing sensitive or classified information from being shared with AI systems. The policy therefore requires staff to follow internal security guidelines when using approved chatbots.