Is that video on your social feed real? Why an Ontario expert now assumes ‘everything is fake’ in 2025
2025 has been the year AI technology has really been taking off with the introduction of higher quality AI-generated videos, filters and more.
By Janis Ramsay // Published December 31, 2025
Artificial intelligence (AI) has advanced leaps and bounds over the last year — and it’s made spotting AI-generated photos and videos online even harder.
As it turns out, you can’t always trust that the heartwarming dog or baby video you’re watching is real anymore.
Ontario-based tech educator and author Avery Swartz said it is getting harder to spot the difference. But a little common sense, and some digging, may help you understand the difference.
AI-generated content has become harder to identify in 2025
This year, AI technology has really advanced with the introduction of higher quality AI-generated photos, videos, filters and more, said Swartz said.
When AI-generated images first hit the scene, it was obvious to many they were computer-generated.
“You would look for things like a person in an AI-generated image might have an extra finger,” Swartz said. “Hands were notorious for looking weird in AI-generated images.”
And for a long time, those images were laughable.
“You had people with seven fingers, their hairline was weird, or a piece of their ear was missing,” she said. “You’d say ‘hahaha, that’s an AI image — I can totally tell.’”
With tech advancements this year, AI images have improved to where everything looks real and you don’t even consider questioning the validity.
And now, these deepfake videos are being used to stir emotions of anger, compassion, and even deceive you into giving scammers some cash.
The technology can also be used to spread lies.
What happens if you go to a resort and don’t enjoy your time? You could manipulate photos to have bugs in the rooms, dirty conditions around the grounds, and post it on social media, Swartz noted.
“You could almost do a smear campaign against the business. If you got a delivery from a pizza place, and didn’t like it, you could totally use AI to add mouldy pepperoni to the pizza and do reputational damage to the business.”
And people could also manipulate photos of relatives or ex-partners they don’t like.
How to identify the AI-generated videos
Sometimes, fake videos — like ones made with Open AI’s Sora — are watermarked with an icon so you can tell it’s been generated, Swartz said.
“Sometimes they are removed or cropped out, so you can’t always assume you’ll know if it’s an AI-generated video.”
Legality of fake videos
In Canada, there is no AI-generation law that says fake videos or images must be disclosed as such.
“It’s kind of like the wild west, it’s totally unregulated for the most part,” Swartz said.
So, the responsibility falls to the viewer — and that’s where critical-thinking skills come into play, she added.
“The younger generations are better at critically thinking about what they’re seeing online.”
Awareness is key
“Odds are, we probably cannot tell it’s AI,” Swartz said. “Here’s the thing — this is a complete flip of how I’ve been most of my life — I now assume everything is fake until it’s proven real.”
Swartz used to think photos were proof of a real-life event, and she pointed out most of the older generations also feel that way.
“I grew up before the internet, and would think ‘oh, a picture, that’s got to be real,’” she said. “Now, I don’t. I assume every image I see is fake. That does weird things to your mind.”
And she thinks that’s where most of us will eventually land.
Just being aware the video you are viewing is likely fake is the first step, Swartz said.
The next step is checking your feelings when seeing videos or images.
“The most convincing ones will drive us to the three emotions — sad, glad or mad. You’ll be sorrowful, it’s controversial or you’ll think it’s funny. Those are the emotions that drive us to share it.”
It’s in those moments you need to pause and ask yourself a few questions, she said.
Does it matter if that AI video is fake?
“Is this actually true? And then, the next question is: Does it matter?”
The real versus fake question might not matter if it’s a video of a cute baby hugging a dog. It’s purely for entertainment value.
But if it’s from a real world news event, or a business smear campaign, then it might matter from an accountability standpoint, Swartz said.
Following Hurricane Melissa, she started seeing fake images and videos of the aftermath. Some videos shared on X showed images of sharks swimming along the streets. It looked believable. But viewers flagged the video, so it now has a note that it is a generated video.
“I try to find the same (video) in the news, ideally from two different news sources,” Swartz said, noting that’s how she attempts to validate it.
Overall, Swartz cautions people to use that critical thinking before believing everything they see.