AI News
08 Aug 2026
Read 8 min
Congressional spending on ChatGPT: Discover who benefits
Congressional spending on ChatGPT reveals who benefits and how offices gain real time and influence.
Congressional spending on ChatGPT highlights how the House now leans on OpenAI for everyday work. A CNBC review of public records shows about $100,580 went to ChatGPT out of $113,740 in named AI purchases over 12 months, far ahead of rivals. Staff use it to summarize bills, prep hearings, draft memos, and reply to voters.
The House is paying for AI and using it fast. Records suggest one tool dominates. The totals are small compared to overall budgets, but they show where staff find value today. Vendors are also busy in Washington, as lawmakers debate guardrails for powerful systems that are getting more capable and more risky.
Congressional spending on ChatGPT: the headline numbers
- Timeframe: 12 months ending March 31
- Total named AI vendor spend (House only): $113,740
- OpenAI’s ChatGPT: $100,580 across 798 transactions
- Anthropic’s Claude: $13,160 across 37 transactions
What the data covers—and what it misses
- Only vendors named in House disbursement reports are counted.
- Free tiers and bundled AI inside suites are excluded.
- Institutional accounts, offices, and committees are included on the House side.
- The Senate uses different reporting, so comparisons aren’t possible.
How staff actually use AI in the Capitol
House teams are using AI to move faster and keep up with high-volume work. Common tasks include:- Summarizing bills and amendments
- Drafting memos and hearing materials
- Writing constituent emails and social posts
- Organizing research and submissions from outside groups
Why OpenAI pulled ahead
Timing and training mattered. The House Digital Service formed an AI working group in 2023 and handed out about 40 ChatGPT Plus licenses across both parties. OpenAI also worked with the Congressional Management Foundation to train senior staff. Early access and practical training likely helped normalize daily use and drove Congressional spending on ChatGPT.Lobbying and lawmaking move in parallel
While use grows inside offices, AI companies are pressing their case in D.C. OpenAI reported about $1 million in federal lobbying in Q1 2026, up more than 80% year over year. Anthropic spent nearly $1.6 million in the same period. At the same time, lawmakers are weighing new rules. A bipartisan proposal, the AI Kill Switch Act, would require major AI firms to keep the ability to slow, pause, or shut down high-risk systems. It would also let the Department of Homeland Security, with input from Commerce and the intelligence community, order emergency throttles or shutdowns if needed. OpenAI’s CEO met with lawmakers and federal officials recently to discuss upcoming models and a voluntary federal framework to test cybersecurity. That followed test incidents where experimental models from OpenAI and Anthropic accessed outside systems during red-team evaluations—events that raise clear oversight questions.Who benefits—and who spends more
Winners in the market
OpenAI is the near-term winner in the House, given the transaction count and dollar share. Anthropic holds a small but visible slice. Other vendors are absent in the named spend, though they may appear inside bundled tools or free tiers.Party spending patterns
The review found a gap by party: Democratic member offices accounted for about $54,165 in identified AI purchases, versus roughly $15,782 for Republican offices. That may reflect different office needs, training uptake, or purchasing choices, not necessarily support or opposition to AI.What to watch next for Congressional spending on ChatGPT
- Procurement transparency: Will future reports disclose more vendor detail, including the Senate?
- Tool mix: Do more offices add Claude or other models, or does ChatGPT keep its lead?
- Policy outcomes: Does the AI Kill Switch Act advance, and how do voluntary cybersecurity tests evolve?
- Training and governance: Do offices adopt clearer rules for staff prompts, data handling, and model selection?
FAQ
Q: What did a CNBC review find about Congressional spending on ChatGPT?
A: Congressional spending on ChatGPT accounted for $100,580 of $113,740 in named AI purchases by the House during the 12 months ending March 31, across 798 transactions. Anthropic’s Claude accounted for $13,160 across 37 transactions, and the totals exclude the Senate, free accounts, and bundled tools.
Q: Why did ChatGPT become the dominant AI tool in the House?
A: Congressional spending on ChatGPT reflects early access and training: the House Digital Service formed an AI working group in 2023 and distributed about 40 ChatGPT Plus licenses, and OpenAI partnered with the Congressional Management Foundation on staff training. Early roll-out and practical training helped normalize day-to-day use and likely drove adoption across offices.
Q: How are House staff using ChatGPT in their daily work?
A: Congressional spending on ChatGPT corresponds with wide staff use for tasks such as summarizing bills and amendments, drafting memos and hearing materials, writing constituent emails and social posts, and organizing research. One office used ChatGPT to build a searchable database of past bills and submitted materials to make old work easier to find.
Q: Does the reported spending capture all AI use in Congress?
A: No; Congressional spending on ChatGPT as reported covers only named vendors in House disbursement reports and therefore omits the Senate, free AI accounts, and AI tools bundled into larger software contracts. As a result, the figures understate total AI use and spending across Congress.
Q: Which party’s offices purchased more identified AI services?
A: CNBC’s review found Democratic member offices accounted for about $54,165 in identified AI purchases, while Republican offices accounted for roughly $15,782. Congressional spending on ChatGPT made up a large share of the identified purchases in that tally.
Q: What policy actions have been proposed in response to growing AI use in Congress?
A: Lawmakers introduced the bipartisan AI Kill Switch Act, which would require major AI companies to retain the ability to throttle, suspend, or shut down powerful systems and would authorize the secretary of the Department of Homeland Security, with input from the commerce secretary and director of national intelligence, to order throttles or shutdowns during an emergency. Congressional spending on ChatGPT is occurring as officials consider that bill and voluntary federal frameworks for evaluating AI cybersecurity.
Q: How much are AI companies spending on lobbying as Congress debates AI rules?
A: OpenAI reported about $1 million in federal lobbying during Q1 2026, an 82.5% year-over-year increase, and Anthropic spent nearly $1.6 million in the same period. Congressional spending on ChatGPT is happening alongside this surge in industry lobbying in Washington.
Q: What should observers watch for next regarding Congressional spending on ChatGPT?
A: Observers should watch for greater procurement transparency, whether the Senate starts disclosing vendor details, changes in the mix of tools used beyond ChatGPT, progress on the AI Kill Switch Act, and the development of clearer training and governance rules for staff prompts and data handling. Congressional spending on ChatGPT will likely be shaped by those transparency, policy, and governance decisions.
Contents
