Mike Judge's Idiocracy has been a punchline for twenty years.
Now a sitting lawmaker just turned the joke into a live demonstration.
A leaked government video caught him reading his own chatbot's notes straight into the official record.
A Career Politician Read His Chatbot's Notes to Itself and Never Noticed
Bill Oliver has sat in New Brunswick's legislature since 2014.
He's served as a cabinet minister and even ran the place as Speaker.
Last month he stood up to address the creation of a new advocate's office and started strong.
"Public confidence in the office of an advocate matters," Oliver told the chamber.
Then something else came out of his mouth.
Oliver then read something a chatbot writes only for the person running the prompt, describing "a more natural, flowing version of that section" it had just drafted for him.
That's not how a politician talks.
It's how a chatbot explains its own edit to whoever asked for it, and Oliver said it out loud without catching the switch.
Nobody in the chamber blinked, and the moment sat unnoticed for weeks.
Then a clip surfaced on the New Brunswick subreddit and Canadians finally got a look at how the sausage gets made.
The video spread fast, and CBC and the Toronto Star both picked it up within days.
An AI-detection service called Pangram later reviewed the speech and rated it highly likely to be machine-written.
https://x.com/Soc_Unveiled/status/2081191091194233044“>https://x.com/Soc_Unveiled/status/2081191091194233044
Reddit users did not go easy on him.
"If he was a lawyer, he'd be sanctioned," one commenter wrote, before asking why an elected official should get a pass a lawyer never would.
Other commenters said the whole thing looked amateurish, and some openly called for Oliver to resign.
A man who spent over a decade building a political career got exposed by nine words he never bothered to read before saying them.
This Is Becoming Every Profession's Problem, Not Just One Lawmaker's
Oliver is far from the first person in a position of trust caught letting a machine do the thinking for him.
Legal researcher Damien Charlotin has tracked roughly 1,800 cases worldwide where AI-hallucinated citations blew up in someone's face, and more than 1,200 of those came out of the United States alone.
Even Latham & Watkins, one of the biggest law firms on earth, got caught submitting a fabricated academic citation generated by a chatbot, and the firm called it "an embarrassing and unintentional mistake."
The Department of Justice has gotten swept up in the same mess.
One case in Mississippi got so embarrassing a judge scrapped the trial entirely after lawyers on both sides submitted fake AI citations.
Bar associations keep threatening discipline and judges keep issuing warnings, and attorneys keep doing it anyway.
Oliver's blunder is funnier because you can watch it happen in real time on video, but it's the same habit spreading through courtrooms, law firms, and now a provincial legislature.
Nobody wants to admit how much of their "own" work is quietly outsourced to a chatbot and never checked before it goes out the door.
Nobody Proofreads Anymore and Everybody's Getting Caught
Here's what makes this one so much fun.
High schoolers get failed and expelled for turning in AI slop without disclosing it, and Oliver just did the exact same thing in a taxpayer-funded chamber and walked out that day without anyone noticing.
He didn't get caught by a fact-checker or a rival party.
He got caught by strangers on Reddit weeks later, watching a video for laughs and realizing the guy on their screen was reading a robot's diary entry out loud.
That's the real punchline: the system built to hold public officials accountable did nothing, and the internet had to do the job instead.
https://x.com/CloseToCredits/status/2081356789514178724“>https://x.com/CloseToCredits/status/2081356789514178724
Twenty years ago Idiocracy imagined a future where nobody in charge could be bothered to think for themselves anymore.
Bill Oliver just proved the movie undersold it, because at least the people in that movie knew they were talking to a machine.
Sources:
- Breitbart, "Canadian Politician Reads AI Writing Instructions Aloud During Legislative Speech," Breitbart, July 27, 2026.
- The Post Millennial, "Canadian Politician Slammed After Reading AI-Generated Text During Legislative Speech," The Post Millennial, July 2026.
- Blaze Media, "Thousands of Lawyers Caught Using AI, and It's Only Getting Worse," Blaze Media, July 2026.
