How to Tell If That Viral Video Is Actually Fake
With AI-powered tools creating lifelike images, videos, and voices at the click of a button, it’s getting harder to trust what we see online. From deepfake political videos to fake music artists with millions of streams, AI-generated content is becoming nearly indistinguishable from the real thing. So how do you know what’s authentic and what’s not?
CTV News spoke with tech analyst Carmi Levy, who shared practical tips for identifying synthetic content—and staying one step ahead of digital deception.
AI Fakes Are Getting More Convincing
You don’t need high-end software to generate a realistic video or voice clip anymore. Apps now let users create AI content with just a text prompt. This rise in accessibility has made fake content explode across platforms.
For example, an indie rock band with over a million monthly Spotify listeners was recently revealed to be entirely AI-generated. In another case, U.S. officials reported attempted impersonation of Secretary of State Marco Rubio using similar tools.
Even CTV News Ottawa wasn’t spared. A fake version of a news segment circulated online, altering anchor voices to falsely promote a scam.
Expert Tip: “Lean In” and Look for Tells
Levy’s best advice? Slow down and lean in. He encourages viewers to actively inspect suspicious content for subtle errors.
“You’ll notice strange movements,” he said. “Things like unnatural motion, lips that don’t sync with the audio, or jerky transitions.”
He also advises watching for lighting issues or odd shadows that seem off. “If someone wraps their arm around another person, the placement might be wrong—too many fingers, or sometimes not enough.”
AI Is Improving—Fast
Unfortunately, these “tells” are getting harder to spot. With every new version of AI software, especially advanced tools like Veo 3, the realism increases.
“These apps are evolving rapidly,” Levy warned. “Even seasoned viewers are finding it difficult to identify what’s AI-created and what’s not.”
Rising Threats and the Canadian Response
The Canadian Anti-Fraud Centre recently issued alerts about scams using AI-generated voice messages impersonating high-profile figures.
In response, the federal government is stepping up. Innovation, Science and Economic Development Canada has created the Canadian Artificial Intelligence Safety Institute, aiming to study risks and protect Canadians from emerging threats.
The government also introduced a Voluntary Code of Conduct in 2023, urging developers to improve AI transparency and provide tools to detect fake content.
How to Stay Safe Online
Levy says it comes down to a basic rule: question everything.
Before you click “share,” take a moment to check the source. Does the content come from a known, trustworthy outlet? If not, it’s best to assume it’s fake until proven otherwise.
“That’s what I do,” Levy said. “I start cynically. I make the content prove it’s real.”
Stay smart, stay skeptical—and stay tuned to Maple News Wire for more updates on tech and security.
