Paragraph 1
A new investigation from CTV News’ Deception Decoded series has uncovered a coordinated network of fake news websites engineered to promote Alberta separatism, amplify disinformation about Canadian politics, and manipulate the answers given by AI chatbots. The sites, which mimic legitimate local news outlets, are not the work of a spontaneous grassroots movement. Instead, the investigation shows they are part of a sophisticated, anonymous digital operation that uses artificial intelligence to generate dozens of articles, then feeds those articles into the information ecosystem that AI assistants such as ChatGPT, Google’s Gemini, and Microsoft’s Copilot rely on. The result is a dangerous feedback loop: a chatbot asked a simple question about Alberta politics can cite a completely fabricated story from a fake news site as though it were established fact. The investigation identifies this as a deliberate tactic to make a fringe political idea appear far more popular and credible than it actually is. By creating a web of interconnected fake sites, all repeating the same talking points about Western alienation, equalization payments, federal overreach, and a supposed imminent breakup of Canada, the operators are attempting to poison the well of public information. The article argues that this is not merely an online nuisance but a form of information warfare that could have real consequences for Canadian democracy, especially in a province where separatist sentiment has historically been volatile and where political leaders have already used threats of secession as a bargaining chip.
Paragraph 2
The scope of the operation is broader than many would expect. The investigation identified dozens of websites, many with names that sound credible and neutral—something like Alberta Times, Western Post, or Edmonton Daily—designed to fool readers into thinking they are established local news brands. In reality, they are what researchers call “pink slime” journalism: low-cost, algorithmically produced content with no real reporters, no editorial oversight, and no transparency about ownership. The articles are often generated using large language models, then lightly edited or republished across the network to create the illusion of widespread coverage. The investigation notes that many of these sites share identical templates, identical privacy policies, and even identical “About Us” pages, which describe fictional newsrooms and nonexistent editorial teams. The content is heavily focused on Alberta separatism, with headlines and stories crafted to stoke resentment against the federal government. Common themes include claims that Alberta is being deliberately impoverished by “Ottawa’s greed,” that the rest of Canada is exploiting the province’s oil and gas wealth, that federal climate policies are a “war” against the West, and that the only solution is full independence. The article also points out that the operation is opportunistic, latching onto real events—such as debates over pension reform, provincial policing, or carbon taxes—and instantly generating AI stories that frame those events as evidence of Canada’s collapse.
Paragraph 3
The most insidious part of the scheme is the way it targets AI chatbots. As the investigation explains, chatbots are trained on enormous amounts of text scraped from the internet. If a fake story appears on multiple sites, with consistent wording and enough search engine optimization, an AI system may treat it as a reliable source. The operators behind the fake news network appear to understand this perfectly. They use search-engine poisoning and interlinking techniques: each fake site links to the others, creating a closed loop of mutually reinforcing disinformation. When someone asks an AI assistant about “Alberta independence” or “Wexit,” the chatbot may pull from these sites and present fabricated events as real. The investigation gives a striking example: a chatbot was asked about recent political developments in Alberta and responded with a detailed summary of a “massive separatist rally” that never happened, based on a story published by one of the fake sites. This is not a glitch; it is the intended effect. The operators are deliberately gaming the AI systems that millions of Canadians now use for everyday information. The report warns that as AI chatbots become more integrated into web search, social media, and even customer service, the ability to corrupt their training data becomes a powerful weapon. The article quotes experts who describe this as a “perfect storm” of cheap generative AI, broken online advertising economics, and the growing public reliance on automated answers.
Paragraph 4
So who is behind it? That is the central question of the investigation, and the answer is complex. Using a combination of domain registration records, hidden tracking codes, server IP addresses, and metadata embedded in the sites, the Deception Decoded team found that the network is not run by a single public figure or a well-known separatist organization. Instead, it appears to be operated by a small group of anonymous individuals with connections to a broader ecosystem of partisan fake-news sites and political influence campaigns. Some of the domains were registered using privacy protection services, but investigators found digital “breadcrumbs” leading to a web of related sites that have been active in Canadian political disinformation for years. The article says there is no direct evidence of a foreign government, such as Russia or China, controlling the operation, but it warns that the tactics are almost identical to those used by known foreign influence networks. The group’s identity remains murky, but the investigation suggests it may be connected to an American-based producer of fake local news content and a Canadian activist with a history of promoting anti-government and anti-establishment narratives. What is clear is that the people behind the sites are not true believers in Alberta separatism. They appear to be driven by a mix of financial profit, through programmatic advertising, and political influence, using a real movement as a vehicle for a broader anti-Canada agenda.
Paragraph 5
The impact of this disinformation goes far beyond the internet. The investigation explains that the fake sites are designed to feed real-world political conversations. Alberta separatism, once a marginal idea, has been pushed into the mainstream in recent years by populist politicians, conservative media, and social media echo chambers. The fake news network amplifies this trend by producing a constant stream of seemingly authoritative content that makes separatism appear inevitable. It can distort public opinion polls, put pressure on politicians, and even influence party leadership races. The report also highlights the damage done to legitimate journalism. Readers who stumble across these fake sites may not realize they are false, and their trust in all media is further eroded. Real Alberta news outlets, already struggling financially, are being undermined by fabricated competitors that look just like them. The article criticizes tech companies for failing to act. AI companies do not disclose their training data, so it is impossible to know how much of their output is influenced by fake content. Social media platforms do not consistently label AI-generated posts. Search engines sometimes rank fake sites above real news. The investigation calls for urgent action: better AI transparency, stronger verification standards, and more public investment in media literacy. It argues that without such measures, the problem will only grow worse as generative AI becomes cheaper and more accessible.
Paragraph 6
In the end, the Deception Decoded investigation is a stark warning about the future of information in Canada. The fake sites boosting Alberta separatism are not an isolated case; they are a test run for a new kind of propaganda that can be deployed anywhere, against any political target. The article closes by noting that even while the investigation was underway, new fake sites continued to appear, and the operators adapted their methods to avoid detection. The challenge is not just to expose one network, but to build systemic defenses against the weaponization of AI. The report makes clear that the threat is not hypothetical. It is happening now, in real time, and the people behind it are anonymous, agile, and largely unaccountable. For Canadians, the lesson is simple: before trusting a news story, or an answer from an AI chatbot, it is necessary to ask who is behind the information. The old rules of media trust no longer apply. In an age when a fake article can be generated in seconds and amplified by algorithms, the line between truth and fiction has become dangerously blurred. As the article concludes, the fight against disinformation is no longer just about fact-checking individual claims—it is about protecting the very foundations of democratic debate. The network behind the Alberta separatism sites may be hidden, but its goal is clear: to make Canadians doubt what is real, distrust legitimate sources, and believe that the only solution is to tear the country apart.