AI Studies

Likely AI Language Surged Across the Congressional Record

At Originality.ai, our analysis of a complete census of 21,388 eligible remarks finds the Likely AI rate rose from 8.4% in 2025 to 12.2% in 2026 YTD.

By Jonathan Gillham, Founder and CEO, Originality.ai

Coverage: January 1, 2025 through August 17, 2026

Originality.ai analyzed every eligible English House and Senate Floor Turn over 100 words in the daily Congressional Record for 2025 and 2026 YTD. 

Of 21,388 turns, 2,099 were classified as Likely AI. The observed rate rose 3.8 percentage points, from 8.4% in 2025 to 12.2% in 2026 YTD.

Why we looked at Congress:

The issue has already drawn national coverage from Politico, which reported that AI-shaped requests were creating problems inside the House Office of Legislative Counsel. Those confidential requests are not public. 

So, we asked a narrower public question: what Likely AI signal is visible in the words attributed to serving members in the Congressional Record?

These are our findings.

Key Findings:

  • The complete census includes 13,338 eligible turns in 2025 and 8,050 in 2026 YTD.
  • The Likely AI rate increased from 8.4% to 12.2%, a gain of 3.8 percentage points.
  • The House rate was higher in both periods: 11.4% versus 3.1% in 2025, and 14.8% versus 6.6% in 2026 YTD.
  • The probable AI signal crossed party lines, but Republican rates were higher within both chambers and years.
  • The census covers 546 politicians. Among members with at least 10 turns, 25 of 270 had a 2026 YTD Likely AI rate above 40%.
Share of Floor Turns Classified Likely AI

The Likely AI Rate Rose From 8.4% to 12.2%

  • In 2025, 1,116 of 13,338 eligible Floor Turns were classified as Likely AI. 
  • In 2026 YTD, 983 of 8,050 were classified as Likely AI. 
  • Every eligible turn was analyzed.
Likely AI Classification by Chamber and Year

House Floor Turns Had the Higher Likely AI Rate in Both Periods

In 2025, 961 of 8,415 House turns were classified as Likely AI, compared with 155 of 4,923 Senate turns. In 2026 YTD, the comparison was 814 of 5,503 House turns, and 169 of 2,547 Senate turns.

Distribution of Likely AI Floor Turn Rates

The Politician-Level Census Widens the Reporting Pool

Among politicians with at least 10 eligible turns, 354 met the threshold in 2025 and 270 in 2026 YTD. In 2026 YTD, 80 had no turns classified as Likely AI, while 25 had rates above 40%. 

Likely AI Classification by Party and Chamber

The Top Observed Politician-Level Likely AI Rate Was 90.0%

Across the combined 2025–2026 YTD period, 495 politicians had more than five eligible Floor Turns. We removed politicians with less than 5 speeches from this analysis. Their observed Likely AI rates ranged from 0% to 90.0%. The chart ranks every qualifying politician from lowest to highest.

Likely AI Floor Turn Rate by Politician

Top 30 Politician-Level Likely AI Floor Turn Rates

Note: The table below excludes the names of the politicians.  

top 30 politician level likely AI floor turn rates

If you are interested in the raw data, please reach out; we are happy to share the data with parties working to further the understanding of probable AI use within government. 

The Likely AI Signal Crossed Party Lines

Likely AI turns appeared among both Democratic and Republican members in each chamber-period. 

In 2026 YTD, Likely AI rates were:

  • 10.8% for House Democrats
  • 19.3% for House Republicans
  • 3.3% for Senate Democrats
  • 9.6% for Senate Republicans

Should Politicians Disclose AI Use?

We found two clear disclosures. 

Senator Marsha Blackburn said she queried Google's Gemma about herself and described the model fabricating an allegation. 

Senator John Cornyn said, “I asked my friendly AI assistant,” before presenting material connected to a request about parole programs. Neither senator said AI drafted the congressional language.

We found no AI-drafting disclosure and no hard AI residue in the U.S. corpus. That differs from the New Brunswick investigation, where apparent drafting instructions were read aloud. 

If AI is used to draft or materially shape words entered into an official legislative record, a simple disclosure would let the public distinguish AI-assisted language from member- or staff-written text.

Final Thoughts

Likely AI language is no longer an isolated occurrence. 

Its sharp growth suggests AI is becoming part of how congressional material is prepared, considering eligible remarks have seen the Likely AI rate rise from 8.4% in 2025 to 12.2% in 2026 YTD.

This shows that probable AI language is materially shaping words entered into the public record.

Methodology

  • Source: GovInfo daily Congressional Record.
  • Scope: every English House or Senate member-attributed Floor Turn strictly over 100 words from 2025 through August 17, 2026.
  • Review: speaker labels were matched to date-specific member terms; 13 former-member or non-speech turns and 34 exact GovInfo overlaps were removed.
  • Detector: Originality.ai with AI Allowance 15% settings; AI confidence of 0.50 or higher was classified Likely AI.
  • Final census: 21,388 turns, all successfully scored. No sampling weights or error bars were used.

How to Read ‘Likely AI’ Classifications

Likely AI is the detector classification for the entire recorded turn. It is not the percentage of a politician’s words written by AI. The classification does not identify who drafted a remark, so individual findings should be reviewed against the official passage and, where important, the member’s office and available video.

September 2026 Study Update

This study has been updated since its initial publication in August 2026. Here’s what’s different from our earlier study:

  • Scope: The earlier study examined CRS summaries, Floor Turns, and Extensions of Remarks from 2020 through 2026. This report narrows the question to Floor Turns from 2025 through 2026 YTD.
  • Coverage: The earlier Floor Turn results came from an outcome-blind probability sample of 5,275 turns. This study scored all 21,388 eligible turns.

AI Disclosure

AI assisted with data processing, quality checks, figure production, and an initial draft. Jonathan designed the study, reviewed and approved this final report.

Jonathan Gillham

Jonathan Gillham

Jonathan Gillham is an engineer, inventor, and entrepreneur. He is the founder and CEO of Originality.ai, an AI content integrity platform that launched the first commercial AI detector in November 2022, just three days before ChatGPT launched. Before founding Originality.ai, Jon worked as an engineer, built and exited two companies. His early work with generative AI in 2020 and 2021 gave him a firsthand view of the coming wave of AI-generated content and the need for technology that could bring transparency and trust to written content. Today, he leads Originality.ai’s work in AI detection and content integrity and is a named inventor on two U.S. patents covering AI detection technology. Jon’s expertise and research have been featured in WIRED, Business Insider, The Register, Global News, The Guardian, Entrepreneur, and The Washington Post, among others.

Al Content Detector & Plagiarism Checker for Marketers and Writers

Use our leading tools to ensure you can hit publish with integrity!

Try our AI Checker now!

cross image
Free Tool Popup image

Sign up now!

Free Tool Image step1
Free Tool Image step2
Free Tool Image step3
Free Tool Image step1
Free Tool Image step2
Free Tool Image step3
Free Tool Image step4
Free Tool Image step5