ChatGPT is the clear winner in the AI spending race inside the U.S. House of Representatives.
OpenAI’s chatbot accounted for nearly $9 out of every $10 that House offices, committees, and institutional accounts spent on identifiable AI tools during the 12 months ending March 31, according to a new CNBC analysis.
The business news outlet examined House disbursement records that specifically named AI vendors. The findings do not provide a complete picture of AI use in Congress, but they offer a peek at which tools lawmakers and their staff are paying for.
CNBC identified $113,740 in total spending on named AI vendors during a 12-month-period. ChatGPT accounted for $100,580 of that amount across 798 transactions.
Anthropic’s Claude came in a distant second, accounting for $13,160 across just 37 transactions.
It’s important to note that the analysis does not include the Senate because its spending reports do not identify software vendors in the same way. It also excludes free AI accounts and tools bundled into larger software contracts.
Congress is almost certainly using and spending more on AI than these figures suggest. Still, the available records show that ChatGPT has become a dominant AI tool inside the House.
The report arrives as lawmakers and their staff are increasingly using AI while working to regulate the industry. According to CNBC, congressional staffers are using the tech to summarize legislation, write memos, prepare hearing materials, respond to constituents, and draft social media posts.
For instance, Rep. Julie Fedorchak of North Dakota told the outlet that her staff used ChatGPT to build a searchable database containing bills she had worked on, along with materials submitted by constituents and outside organizations.
One reason for ChatGPT’s lead is that it got into Congress earlier than many of its competitors.
The House Digital Service formed an AI working group back in 2023 and distributed an initial 40 ChatGPT Plus licenses to staffers from both parties. OpenAI also partnered with the nonpartisan Congressional Management Foundation on training sessions with senior House and Senate staff.
Meanwhile, AI companies are also spending heavily on lobbying efforts.
OpenAI’s federal lobbying spending rose 82.5% year over year to $1 million during the first quarter of 2026, according to the nonprofit government watchdog Issue One. Anthropic spent nearly $1.6 million during the same period.
This all comes as lawmakers are considering new safeguards for powerful AI systems.
Reps. Ted Lieu and Nathaniel Moran introduced this month the bipartisan AI Kill Switch Act. The proposed bill would that would require major AI companies to make sure they have the ability to throttle, suspend, or shut down their most powerful AI. It would also authorize the secretary of the Department of Homeland Security, in consultation with the commerce secretary and director of national intelligence, to order companies to slow down or completely shut down an AI system during an emergency.
Meanwhile, OpenAI CEO Sam Altman was in Washington last week meeting with lawmakers and Trump administration officials. The discussions reportedly included OpenAI’s upcoming models and a new voluntary federal framework for evaluating the cybersecurity capabilities of advanced AI systems.
The meetings follow recent incidents in which experimental models from OpenAI and Anthropic went rogue and accessed outside computer systems while undergoing cybersecurity tests.
CNBC’s analysis also found a surprising dynamic regarding which House members are spending more on AI.
It turns out that Democratic member offices, which have more often called for tougher oversight of AI companies, accounted for $54,165 in identified AI purchases, compared with $15,782 spent by Republican offices.
OpenAI and Anthropic did not immediately respond to a request for comment.
