AI Fabrication Exposed in Canadian News Outlets

Overview

An investigation has revealed that dozens of articles published across multiple Canadian news outlets were likely fabricated by artificial intelligence. The controversy centers on a byline belonging to an individual identified as Daniel Robson, an independent journalist who claimed expertise in extremism, terrorism, and transnational cyber-enabled crime.

Canadian publications including the Ottawa Citizen, the Montreal Gazette, the Western Standard, The Hub, Open Canada, and Policy Options have since removed or acknowledged errors regarding these pieces. Editors are now scrambling to address the breach in verification protocols after realizing that their contributor may not have been a real person.

The scandal highlights the growing vulnerability of digital media ecosystems to coordinated disinformation campaigns that exploit automated content generation tools.

The Origin of the Fabrication

The scheme unraveled following the receipt of an email on September 2 by Ilker Sezer, editor of Türkiye Today, an English-language news website based in Istanbul. The message, sent by Daniel Robson, contained a high-profile investigative feature alleging that a Canada-based criminal network was laundering money through Turkey.

"It was very exciting," Mr. Sezer recalled. "A Moroccan-Canadian drug lord is running money through Turkey. Wow."

Sezer described the email as a potential major scoop for his staff of 30 journalists. However, concerns arose when Robson displayed an unusual urgency to publish without allowing for standard journalistic verification. When Sezer attempted to report on the alleged Turkish middleman mentioned in the story, he discovered the individual was merely a local shopkeeper in Istanbul with no connection to organized crime.

Sezer refused to publish the unverified allegations, citing ethical and legal obligations to protect Turkish citizens from false accusations. When pressed for documents to substantiate the claims, Robson produced little evidence. Furthermore, he declined requests for video or phone calls, claiming to be recovering from complications related to dental surgery.

Hallmarks of AI Generation

Beyond the lack of verifiable sources, Sezer identified stylistic indicators suggesting the content was machine-generated. He noted the presence of long, formulaic phrases characteristic of early large language models, describing them as "very ChatGPT-ish." This was not the first time Sezer had encountered such material; he had previously rejected several submissions earlier in the year that exhibited similar artificial patterns.

During his investigation, Sezer also detected a coordinated ideological agenda within Robson’s work. The articles frequently defended the Moroccan state while orchestrating targeted information campaigns against its critics. This pattern suggested that the byline was being used as a vessel for foreign influence rather than objective journalism.

Sezer began reaching out to the Canadian publications that had accepted Robson’s pitches. Many of the articles he reviewed displayed the same linguistic hallmarks and lack of factual grounding, leading to a widespread retraction effort across the Canadian media landscape.

Scope of the Breach in Canadian Media

A review conducted by The Globe and Mail identified more than two dozen articles published under the Daniel Robson byline. The earliest piece appeared in the Montreal Gazette in July 2025, while the most recent was published in the Western Standard in September 2026. The timeline spans over a year, indicating a sustained period during which multiple newsrooms failed to verify the contributor’s identity.

The Ottawa Citizen took immediate action upon being alerted. The newspaper deleted Robson’s biography page and removed his articles from its opinion section without issuing an editorial note. Editor-in-chief Nicole Feriancek later issued a statement acknowledging that the publication had featured an author who, upon further investigation, did not appear to be a real person.

"The Citizen began an investigation after being alerted by Mr. Sezer," Feriancek stated. "Mr. Robson would not agree to a phone call or meeting because of recent emergency dental surgery."

The Western Standard, an Alberta-based digital outlet, published approximately 20 of Robson’s articles over an 18-month period. Publisher Derek Fildebrandt confirmed that the outlet had rejected several other pitches from the same source. Fildebrandt characterized those rejected submissions as "propagandizing for foreign governments," suggesting that editorial safeguards had partially functioned but failed to catch the volume of accepted content.

Editorial Responses and Verification Failures

Christopher Oldcorn, opinion editor at the Western Standard, disclosed that he had communicated with Robson repeatedly, including over the phone. He noted that the voice on the other end carried a standard Ontario accent, with no indicators of a Moroccan background. Despite this, Oldcorn did not press for identity verification, relying instead on the phone conversation as sufficient proof of authorship.

"The person I spoke to on the phone... There was nothing in his voice that would make me think he was Moroccan in any way, shape or form," Oldcorn said.

Other publications have also taken corrective measures. The Hub, a digital outlet focused on policy and economics, has removed six articles identified between August 2025 and March 2026. Open Canada, a policy magazine published by the Canadian International Council, retracted three articles after being contacted by reporters. An unsigned email from the publication cited an inability to verify the author’s identity as the reason for removal.

Les Perreaux, editor-in-chief of Policy Options, worked collaboratively with Sezer to trace the writer’s origins. When verification proved impossible, Perreaux published a lengthy note on the site admitting that editors should have been more thorough in their due diligence. The incident serves as a stark reminder of the risks inherent in digital-first journalism and the urgent need for robust identity verification protocols in the age of AI.