OPINION: The CRTC needs a plan for fake public opinion

One of the people who supposedly weighed in on America's net neutrality rules in 2017 had been dead for seven years. Someone used a New York man's name and address to urge the Federal Communications Commission to repeal the rules. He had, of course, expressed no such view. By the time the proceeding closed, the FCC had received more than 22 million comments. Its previous record was 3.7 million. In 2021, the New York attorney general reported that nearly 18 million were fake. The fraud came from both sides of the argument. Broadband for America, an industry group, spent $4.2 million on a campaign that produced more than 8.5 million fake FCC comments and over half a million fabricated letters to Congress. On the other side, a 19-year-old California student submitted 7.7 million comments supporting net neutrality. Another 1.6 million came from a source investigators could not identify. The firms involved had done this elsewhere, too. Three lead generators had worked on more than a hundred other campaigns aimed at regulators and public officials, producing over a million additional fake comments and 3.5 million fabricated digital signatures. Put that $4.2 million beside the 8.5 million FCC comments and the cost was roughly fifty cents apiece, with the congressional letters on top. Much of the work was crude. Researchers spotted hundreds of thousands of identical submissions, and people found themselves and dead relatives in the record. Some contractors were already using software to vary the wording. That was nine years ago. What concerns me now is how little a fabricated intervention needs to stand out. An AI system can write a plausible submission in the voice of a business owner, vary the argument and put it in the right format. It can even make it appropriately boring. There need be no repeated paragraph for a researcher to spot. Writing the text is only part of running a fraudulent campaign. Identities, submission systems and scrutiny still present obstacles. But the work of producing thousands of different, readable comments has become much easier. Looking for matching sentences will not be enough. Canada has reason to pay attention. The CRTC accepts written interventions from interested members of the public, as it should. People affected by a telecom decision ought to be able to tell the regulator what it means for them without first hiring counsel. That access is worth protecting. The Commission also has safeguards already. Its procedural rules let it require verification by affidavit and demand evidence that a representative has authority to speak for someone else. Those are useful powers. I went looking, though, for a published policy explaining how it handles AI-assisted submissions and suspected mass fabrication. I did not find one that answered those questions. If I missed it, I would welcome the correction. This matters because submissions become part of a record on which decisions are made and defended. Telecom decisions can be challenged through petitions to Cabinet or, with permission and on questions of law or jurisdiction, appeals to the Federal Court of Appeal. A fake comment would not automatically undo a decision. Reliance on fabricated evidence could, however, give a losing party something serious to investigate. The Commission should be able to explain what it checked and why it trusted the material it used. There are some practical things it could do now. Start by explaining how submissions are weighed. If a thousand versions of the same argument add little beyond the first, say so clearly in the guidance participants read before filing. Equally, a short account from an actual customer may contain something the industry's lawyers have missed. What matters is what the submission contributes and whether there is a real person or organization behind it. Then publish a policy on AI-assisted drafting. A ban would be difficult to enforce and would sweep in ordinary editing and accessibility tools. Require disclosure of substantive AI use, define what that means, and explain how it will be treated. Someone who approves an AI-assisted submission should remain responsible for its facts and arguments. Disclosure alone should not make their views count for less. Finally, keep responsibility for the record inside the Commission. Research has shown weaknesses in AI text detection, including its vulnerability to paraphrasing. More fundamentally, a detector cannot establish whether the named person authorized a submission. I sell AI software for a living. I would not buy a detector to settle that question. The useful checks concern consent, identity and evidence. Who approved the intervention? Can the person filing it substantiate its claims? If a campaign organizer says it represents ten thousand people, what supports that assertion? The Commission already has procedural tools for asking these questions. It should explain when it will use them. The 2017 fraud was carried out by marketing firms doing familiar commercial work. Canadian telecom proceedings have enough money riding on them to make manufactured support an obvious concern. The next campaign may arrive as a collection of quite ordinary submissions. The CRTC should have its answer ready before it has to defend a decision built on them. Al Vigier is the CEO of Caseway, an AI company that co-creates products with large enterprises and shares the revenue with them. He spent seven years in the Canadian Army, where he was shot. The positions and opinions expressed by The Wire Report columnists and contributors are solely reflective of the author’s perspective and do not necessarily reflect The Wire Report’s views or values.

OPINION: The CRTC needs a plan for fake public opinion

Opinion |
By Al Vigier
|September 29, 2026

Unlock all the Canadian telecom, broadcasting and digital media news you need.

FREE TRIAL

SUBSCRIBE

Inside The Wire Report

Information

Hill Times Publishing

Get in Touch With Us

OPINION: The CRTC needs a plan for fake public opinion

Opinion | By Al Vigier |September 29, 2026

Unlock all the Canadian telecom, broadcasting and digital media news you need.

FREE TRIAL

SUBSCRIBE

CRTC approves sixth area code for southern Alberta

Liberals accused of violating CRTC independence in regulatory challenge

CRTC throws out Videotron Part 1 over Ottawa transmission lines

Related Stories

Solomon addresses AI concerns as protester hits stage at conference

Bengio says Canada, Germany investment will support LawZero proof of concept

Bell to quadruple AI data centre in Saskatchewan

Doug Ford taps new telecoms lead in cabinet, creates AI adoption portfolio

Reuse & Permissions

Inside The Wire Report

Information

Hill Times Publishing

Get in Touch With Us

Opinion |
By Al Vigier
|September 29, 2026