A majority of the American public wants Congress to regulate AI. But as policymakers try to make informed decisions, Senate lawmakers and their staff are barred from using all but the most basic AI tools.
According to an internal communication obtained by NPR that was sent to Senate staff in the spring by the sergeant at arms, who oversees cybersecurity in the chamber, Senate staff have access to three AI chat interfaces at no cost to their offices: Microsoft Copilot Chat, Gemini Chat for Google Workspace Enterprise Plus and OpenAI ChatGPT Enterprise.
But in an apparent effort to balance innovation with risk, the sergeant at arms has not currently authorized any of the more advanced AI tools that are at the center of regulatory discussions in Washington, as well as recent public debate about AI development and safety. That debate has been prompted in part by AI industry leaders' own warnings.
These more capable tools, like OpenAI's Codex or Anthropic's Claude Code and Cowork, allow "agents" powered by advanced AI models to complete tasks autonomously at a user's request — and can be powered by models of varying cost and capabilities.
"Congress, unfortunately, has always lagged behind," said Adam Kovacevich, the founder and CEO of center-left tech industry policy group Chamber of Progress. "Right now you've got lawmakers writing rules for technology they, in many cases, never even used and that's a problem."
The agentic tools, which can be given varying levels of access to a computer's files, are considered more capable than the web-based chat interfaces that Senate staff currently have access to. The rules from the sergeant at arms state that "platforms cannot independently access internal Senate drives, shared folders, email, Teams chats, or other Senate resources."
While outside experts acknowledge there are risks of a broader embrace of AI tools by Congress, they say limiting access also has its drawbacks.
"The use of tools and technologies that have the possibility of exfiltrating data elsewhere are a significant risk for the Senate and for the House and elsewhere," said Daniel Schuman, executive director of the American Governance Institute, a nonpartisan think tank focused on strengthening government institutions. "This is also true with even more prosaic tools, right — the use of Google documents or email."
"But those concerns are not the best reason for the non-approval of certain technologies," he said. "I think people want to do better, but the incentives of the way the institution is set up is a drag on progress."
Advanced AI tools are being vetted for specific use cases, according to a Senate staffer who spoke on the condition of anonymity because they were not authorized to publicly discuss the chamber's cybersecurity policies. No timeline was offered for that work.
"I think security is often a pretext for slower government agencies and businesses to use," said Kovacevich. "The reality is that you've got Fortune 500 firms who have plenty of sensitive customer information and data using all these tools and using them successfully. So I tend to think those arguments are sort of pretextual objections."
The top Republican and Democrat on the Senate Committee on Rules and Administration, which oversees the Senate sergeant at arms, did not address specific details about the policy in response to questions from NPR about the chamber's AI rules. But in a statement, a committee spokesperson said the policies are designed to "protect Senate data and include stronger security and contractual requirements."
"While Senate offices should have access to AI tools, subject to individual members and committee policies, it's critical that these powerful tools are adopted with adequate safety and security checks to ensure that the Senate's constitutionally protected data is secure," the statement said.
Across the Capitol in the House, AI use is overseen by the Office of the Chief Administrative Officer, a separate body. A House staffer who spoke on the condition of anonymity because they were not authorized to publicly discuss chamber cybersecurity policy said that House rules outline specific use cases in which AI use is permissible and said tools from Microsoft, OpenAI, Google and Anthropic are approved in various scenarios.
The CAO office did not immediately reply to a request for comment.
Finding a path forward
Schuman suggested that if Senate leaders decided there was a need to access more powerful AI technologies, they could make it happen.
"Their senior staff can have conversations and say, 'we really want you to expedite this, we think that there's value here and you should do it.' And the Sergeant at Arms is probably going to be responsive to that," he said. "But it's also worth keeping in mind, this is a political institution."
The sergeant at arms "is going to be more reticent because they don't want to have the finger pointed at them in case things go wrong, so they have an incentive to put the brakes on most things," Schuman continued.
Meanwhile, it seems some offices are moving ahead with these tools on an ad hoc basis. Sen. Bernie Moreno, Republican of Ohio, spoke to technology news show MTS Live last week about how his office uses AI. He said his office has an AI staffer they named Albert.
"He works long hours, doesn't require a lot of pay, but it's great for research, things that would normally take somebody a long period of time to kind of vet," Moreno said. "We use it quite a bit. I personally would like to be using it a lot more, but unfortunately the Senate has some crazy rules about how we're allowed to deploy or not allowed to deploy."
"Right now, the Senate prohibits us from integrating outside systems into the Senate's," Moreno said, "but I'm hoping that, soon in fact, that they get that done next year."
Moreno's office did not respond to an inquiry about how Albert is powered and whether it complies with Senate rules.
"This is a tool that half of all Americans are using today," Kovacevich said about AI chatbots, criticizing lawmakers who refused to use the technology they are being called upon to regulate — including one senior Democrat who told Axios last week that "you don't need to do drugs like heroin to know they should be regulated."
"Well, that's true, but nobody thinks heroin has an upside to be preserved," Kovacevich said. "So if your only mental model of the technology is the harm, you're probably going to write a bill that's all brakes and no steering wheel."
Copyright 2026 NPR
