AGP Picks
View all

A Politician’s On-Air Slip Exposed a Larger Pattern of 'AI Residue'

AI Residue

Originality.ai detected traces of AI-generated content in 22.9% of eligible New Brunswick legislative turns.

These traces are especially important in public communications because they raise questions about review, disclosure and accountability.”
— Jon Gillham
COLLINGWOOD, CANADA, July 30, 2026 /EINPresswire.com/ -- Originality.ai has published a new study examining “AI residue” ; the unintended instructions, drafting notes, and other traces of AI assistance that remain in finished content.

The study follows an analysis of a June 9, 2026 speech by New Brunswick politician Bill Oliver during debate on Bill 46. In the official legislative video, Oliver appeared to read drafting language aloud, including: "Here's a more natural flowing version of that section."

Originality.ai transcribed the approximately 34-minute speech and divided it into 24 timestamped blocks. The full transcript and 19 of the 24 blocks were classified as Likely AI using Originality.ai with a 15% AI Allowance.

A Likely AI classification does not prove that AI wrote a speech. Detector confidence scores also do not represent the percentage of words written by AI.

BROADER ANALYSIS FINDS RISING LIKELY AI CLASSIFICATIONS

The incident prompted Originality.ai to examine 204 official New Brunswick Hansard transcripts covering March 2022 through May 13, 2026. The dataset contained 18,023 pages and more than five million words.

After limiting the detector analysis to English-language speaking turns longer than 100 words, researchers evaluated 8,175 speaking turns from the official record.

Key findings include:

• Over the latest 12 months, 499 of 2,181 eligible speaking turns, or 22.9%, were classified as Likely AI.
• Likely AI classification rates increased each year included in the analysis.
• Among 28 politicians with at least 30 eligible speaking turns, 14 had classification rates of 25% or higher and five had rates of at least 50%.
• Bill Oliver's rate was close to the group average. Nine of his 34 eligible turns were classified as Likely AI, compared with 26.1% across the 28 qualifying politicians.
• Statements of Condolence and Congratulations had the highest classification rate at 59.3%, compared with 25.3% for Statements by Members and 21.5% for Statements by Ministers.

Oliver's June 9 speech was outside the broader analysis window because its official transcript was not yet available at the study cutoff.

"AI Residue gives us a useful name for the drafting instructions, placeholders and other AI fingerprints that people unintentionally leave behind," said Jon Gillham, founder and CEO of Originality.ai. "These traces are especially important in public communications because they raise questions about review, disclosure and accountability."

METHODOLOGY

Researchers collected 204 official New Brunswick Hansard transcripts and retained the original floor-language column. Consecutive remarks by the same politician were combined when no other identified speaker intervened.

This produced 17,582 speaking turns from 76 politicians. The detector analysis included 8,175 English-language turns longer than 100 words. Each complete turn was scanned once using Originality.ai Lite with a 15% AI Allowance. Scores of 50% or higher were classified as Likely AI.

Hansard is an edited official record rather than a verbatim transcript. The findings should therefore be interpreted as detector classifications within that record, not proof that individual politicians used AI to write their remarks.

SUPPORTING MATERIALS

Full study: https://originality.ai/blog/politician-reading-ai-residue-reveals-a-larger-ai-pattern
Official Legislative Assembly webcasts: https://www.legnb.ca/en/webcasts
June 9 video excerpt: https://youtube.com/shorts/16WYL2Wg6sk?feature=share
Originality.ai analysis of the full speech: https://app.originality.ai/share/xiwzrdp6jmgl3y05

ABOUT ORIGINALITY.AI

Originality.ai provides AI detection, plagiarism checking, fact checking and content-quality tools for publishers, agencies, educators and content teams. Its research examines how generative AI is changing online information, professional communications and public-facing content.

Madeleine Lambert
Originality.ai
maddie@originality.ai

Legal Disclaimer:

EIN Presswire provides this news content "as is" without warranty of any kind. We do not accept any responsibility or liability for the accuracy, content, images, videos, licenses, completeness, legality, or reliability of the information contained in this article. If you have any complaints or copyright issues related to this article, kindly contact the author above.

Share this page:

Advanced Search Options

Search for:

Search scope:

Type:

Search in:

Date range:

The last

Sort by:

Sign up for:

Political Journal: Canada

The daily local news briefing you can trust. Every day. Subscribe now.

By signing up, you agree to our Terms & Conditions.