Imagine scrolling through X on a quiet afternoon, and suddenly there he is: Jagmeet Singh, the former NDP leader, being handcuffed by police. Another image shows him clutching a suspicious duffel bag, and another is something straight out of a crime drama—a mugshot. It feels alarming, even real for a split second. But none of it happened. These are AI-generated fabrications, part of a growing wave of deceptive advertisements designed not to inform you, but to steal from you. In just 30 seconds of casual scrolling, reporters discovered four separate versions of the same fake story fluttering through the platform’s feed. Singh is only the latest public figure to be dragged into this unsettling trend; the ads use his face because it is recognizable, because it makes people stop, and because outrage and curiosity are powerful clickbait. The actual goal is not political commentary. It is financial fraud disguised as breaking news. The fake images are deliberately outrageous, engineered to bypass your skeptical instincts and get you to click before you pause. Once you click, however, you leave the realm of manipulated photographs and enter a labyrinth of false journalism, high-pressure investment pitches, and potential identity theft. This is the new face of online scams in the age of generative AI: cheap, fast, visually convincing, and disturbingly personalized to exploit the public trust we place in politicians and news outlets. And according to experts, it is only going to get harder to tell what is real. While deceptive political ads are nothing new, the speed and scale at which they can now be produced have changed the game entirely. What used to require a Photoshop expert, a printing press, or a gullible audience can now be generated in seconds by anyone with an internet connection and a grudge against accountability. The result is a digital ecosystem where a loved journalist, a former party leader, or a fictional news brand can all be weaponized against ordinary people who just want to stay informed. That is exactly what is happening to Singh, and, as cyber investigators are discovering, to countless Canadians who have no idea they are being herded toward a very well-disguised trap.
Once you click on these ads, the deception deepens. Most of them lead to a fabricated CBC News article that alleges Singh was arrested after appearing on CBC’s Power & Politics to promote an AI-powered investment platform called Groman Pro. The story claims the platform allows Canadians to earn passive income through automated AI trading, with a minimum deposit of $350. It goes further, inventing an on-air confrontation between Singh and former housing minister Sean Fraser, followed by Singh’s arrest outside CBC’s Ottawa studios. It says more than 47,000 Canadians registered in five hours and promises monthly earnings between $1,200 and $2,800. None of this is true. The article provides no verifiable evidence, no court records, no police statements, no reputable financial analysis. Instead, it ends with a countdown-style warning, urging readers to register immediately before availability closes and deposit their money. This is a classic urgency tactic: create a false deadline, make the offer seem exclusive, and pressure people into acting on emotion rather than logic. Some ads don’t even bother with the fake article; they send users straight to an AI chatbot that poses as a representative of Groman Pro, ready to answer questions and nudge people toward sending money. Independent cybersecurity experts at Gridinsoft, a Canadian-Ukrainian company, examined Groman Pro and found a pile of red flags: a low third-party reputation score, high-risk trading behavior, patterns consistent with financial scams, almost no independent reputation data, and a domain that was only 12 days old. As Gridinsoft noted, common bait includes “guaranteed” returns, urgent deposit windows, and account managers pushing larger transfers. In many cases, users can easily deposit money but then find themselves unable to withdraw without paying extra “verification” or “tax” fees. The whole operation is engineered to take money and sensitive personal information, not to generate returns. And by dressing the scam in the visual language of legitimate news—the CBC logo, a recognizable politician, a talk-show setting—it feels far more trustworthy than a random late-night pop-up. That trust is precisely what the scammers are selling.
Vaseem Baig, an AI educator known online as Vaseem the AI Guy, has spent time studying these deceptive campaigns. He calls the Singh ads “fraud dressed up as political news.” The arrest narrative, he says, is the hook—a shocking, emotionally charged story that makes people click before they stop to consider whether it is real. Political figures make especially effective bait because they already live in the public imagination. People have strong feelings about them, and those feelings can be tapped in an instant. “Politicians are now being used because public outrage and curiosity—it’s like clickbait, right? [It] just gets people going,” Baig explained. But the underlying scam is not new. Long before generative AI, fraudsters ran fake investment schemes and wrote fake celebrity endorsements. What AI changes is the economics of deception. “AI can invent scams, but AI does not invent the scamming mentality,” Baig said. “It just makes scammers cheaper and faster and more convincing at scale.” In the past, a fake news article might have obvious spelling errors, awkward layouts, or low-quality images. Today, AI-generated content is polished enough that most people cannot identify it by looking for visual mistakes or poor writing. The use of familiar news brands like CBC adds another layer of credibility, making it even harder for readers to notice that they are leaving reality behind. Baig warns that this erosion of trust cuts both ways. Every fake arrest story that circulates, even if it is eventually debunked, chips away at the credibility of real journalism. People begin to wonder whether any news is trustworthy, and that cynicism undermines the entire information ecosystem. He also noted that when bad actors use AI in this way, it damages public confidence in AI technology itself. Tools that could be used for education, creativity, and productivity instead become associated with fraud and manipulation. The result is a world where we are all a little less certain about what is true, what is false, and who is really behind the image on our screen. That uncertainty is a feature, not a bug, of modern scams. It keeps us hesitant, and hesitation can make us more likely to trust a fake that feels familiar rather than investigate the unfamiliar truth. Baig’s message, ultimately, is that the technology is not the villain—the people wielding it are. But the technology has handed them a megaphone, and we are only beginning to understand how loud it can get.
So what is being done about this flood of fake ads? On X, many users say they have noticed an increase in AI-generated ads, with little evidence that the platform is removing them or notifying those who have been targeted. Now Toronto reached out to X for comment but did not hear back by publication. Baig argues that platforms like X have both a moral and practical responsibility to know what is running on their networks. He suggests they should verify advertisers before accepting payment, especially when the ads feature recognizable people or well-known news brands. He also recommends automated tools designed to detect impersonation campaigns, and clearer escalation procedures for users who report suspicious content and then hear nothing back. At the legislative level, MP Melissa Lantsman has been trying to push things forward with the proposed Protection Against Online Fraud Act. Lantsman says the issue first became glaringly visible about 18 months ago, when seniors started contacting her about similar scams. Those earlier versions of the fraud did not feature Singh; instead, they showed fake articles with CBC senior investigative reporter Erica Johnson covered in green slime, supposedly leaving a TD bank, as part of another investment hoax. That ad has since disappeared, replaced by the Singh versions, but the pattern is identical. Lantsman describes these scams as highly predatory. They don’t only steal money; they steal autonomy and independence. They target people who may not be as familiar with the internet, people who are looking for financial security and who are vulnerable to suggestions of guaranteed returns. By the time victims realize they have been tricked, their savings may be gone, their personal information may be in the hands of criminals, and the trail is often impossible to follow. Lantsman’s frustration with tech companies is clear. She believes billion-dollar platforms have been allowed to act as unwitting—and sometimes uncaring—accomplices to this fraud. The lack of enforcement and accountability means the scammers face almost no consequences, while ordinary people bear the financial and emotional brunt. That, she says, cannot continue.
The Protection Against Online Fraud Act is Lantsman’s attempt to change the calculus. If passed, it would require large online platforms to take meaningful steps to prevent and remove fraudulent content before it reaches users. It would hold companies accountable for the harmful material hosted on their platforms, forcing them to proactively identify scams and take them down rather than waiting for complaints. It would also require platforms to notify people who have been targeted, and to communicate clearly about what a user clicked on when it turns out to be fraudulent. In that way, the bill tries to shift the burden from individual users to the corporations that profit from their attention. Lantsman says the bill also includes a new section for the criminal code, one that would require judges to consider factors like age, health, and financial vulnerability as aggravating circumstances when sentencing online scammers. This means that scamming a senior living on a fixed income, or someone facing illness, could be treated as a more serious crime than the same fraud against a financially stable younger adult. It is a small but important recognition that online fraud is not a victimless crime, and that the consequences can be devastating in ways that go beyond the immediate loss of money. The bill is not law yet; Lantsman says the House of Commons will likely hear it in spring 2027. That timeline may feel painfully slow to the people who have already been hurt by these schemes. But Lantsman stresses that the actual government has the opportunity to do something about it now, and that the proposed bill gives them a clear path forward. She is blunt about the current state of affairs: “The problem is that nobody gets in trouble for the slop.” Without legal consequences for the platforms that host the slop, or for the scammers who create it, the ads will keep coming. The Singh images are simply one iteration of a business model that works because the risks are low and the rewards are high. The bill aims to raise those risks and lower those rewards, but until it passes, Canadians are left to protect themselves with little more than awareness and skepticism.
In the meantime, both Lantsman and Baig want Canadians to know that they are not powerless. There are practical steps that can help people avoid getting caught in these AI traps. Baig advises paying close attention to website addresses rather than relying on familiar logos. Logos can be copied or recreated with astonishing ease, and scammers can make a fake news page look almost identical to the real one. He also warns against sending money to any AI-powered trading platform that has not been independently verified. The promise of guaranteed returns is, by definition, a red flag; legitimate investments always carry risk. Anyone who sees one of these ads should report it to the platform, and should also talk about it with friends and family. Awareness is a form of protection, especially for people who may not be as comfortable with technology. Lantsman echoes that sentiment, adding that the emotional toll of these scams is just as important as the financial one. She has heard from victims who were embarrassed, ashamed, and afraid to tell anyone what had happened. Humanizing the issue means recognizing that behind every click is a person—a person who may be struggling, who may be trusting, and who may be trying to find a way to make ends meet. Scammers exploit that humanity, and the best defense is to build a culture where people can talk about these experiences without shame, where reporting a suspicious ad is a normal act, and where skeptical curiosity is encouraged. The fight against AI-driven fraud is not just about code and policy; it is about preserving the basic ability to believe what we see, to trust what we read, and to know that the people on our screens are who they say they are. As AI becomes more sophisticated, that trust will become more fragile. But it is not impossible to protect. By staying informed, by verifying sources, and by refusing to let urgency override judgment, Canadians can reduce the chances of falling victim to the next scheme. The fake images of Jagmeet Singh will eventually disappear, but they will be replaced by something else, because that is how the scam industry works. The only real defense is a public that understands the bait, recognizes the hook, and refuses to bite.