How many Canadian celebrities like TV chef Mary Berg, crooner Michael Bublé, comedian Rick Mercer, and hockey megastar Sidney Crosby are finally revealing their secrets to financial success? Is that surprising? That was until the Bank of Canada tried to stop them.
None of it is true, of course, but it's a metaphorical bag of magic beans that obvious scammers on social media have tried to sell to people, enticing users to click on sensational posts and He was arrested and Mr. Bublé was taken away. On his website for CTV News, at first glance it appears to be a legitimate news article.
If you are further intrigued by the supposedly AI-generated article, there is a good chance that you will click on a number of links (about 225 per page) directing you to sign up and hand over your first investment. It is said that $350 will increase more than 10 times in just 7 days.
These were just the latest in a series of deepfake ads, articles, and videos that exploited names, images, footage, and even audio of prominent Canadians to promote investments and cryptocurrency schemes.
Lawyers with expertise in deepfakes and AI-generated content warn that there are currently few legal recourses and Canadian law is not evolving as rapidly as the technology itself.
Financial fraud and schemes to steal celebrities' likenesses are nothing new, but the use of rapidly advancing generative AI technology “puts a new spin on a pretty old concept,” says lawyer and partner at Toronto-based Triz LLP. says Molly Reynolds.
And it will get worse before it gets better. Developing tools and laws to prevent it is a catch-up game we're already starting to lose, she said.
Information Morning – NS7:01Impact of deepfake fraudulent advertising
If you spend any time online, you've probably seen it. Ads that show CBC hosts and talents like Elon Musk paying money in some kind of get-rich-quick scheme. These fall into the category of “deepfakes,” or videos generated by AI. Our technology columnist Noor Jinshir Heywood weighs in on this.
Deepfake detection
There's a lot of content on the internet that has obvious signs of AI generation, but WonSook Lee, a computer science professor at the University of Ottawa, says some of it is so good that it's now much easier to tell what's real. said that it has become difficult.
She said that even a few years ago, AI-generated images of people or deepfake videos could be easily detected by looking at them and noticing differences in pixelation or composition. But some programs now allow you to create near-perfect photos and videos.
Anything that isn't produced perfectly can be further modified using photo or video editing software, she added.
As we learn more about AI, it also gets smarter.
“If we can find a way to detect deepfakes, we can help improve deepfakes,” she said.
Scammers are using deepfakes of trusted celebrities to scam people out of their money through fake online ads. National's Ian Hanomansingh is one of them. He looked at what the law says and what social media companies are doing about it.
star power
It appears that X has somewhat reined in the swarm of Canadian celebrity scam ads, suspending some (but not all) of the accounts that share them. CBC News attempted to contact a spokesperson for the social media platform's parent company, X Corporation, but received only an automated response.
X and other social media and website hosting companies may have policies aimed at preventing spam and financial fraud on their platforms. But Reynolds said they faced “questions of moral obligation and legal obligation.”
That's because platforms have fewer legal obligations to remove fraudulent content, she says.
“There are people who are severely affected, with no legal recourse, no help from technology companies, and probably no large social networks like Taylor Swift to rely on,” Reynolds said.
After all, no famous Canadian has had as much influence as Taylor Swift. If so, perhaps the story would play out differently.
rapid spread AI-generated sexual images Social media companies sprung into action almost immediately following the pop music superstar's comments last month. The White House also weighed in.
X quickly removed the image and blocked searches for Swift's name.Within days, U.S. lawmakers submitted an invoice To fight deepfake porn like this.
U.S. White House Press Secretary Karine Jean-Pierre responded to a reporter's question about a blatant fake image of Taylor Swift generated by artificial intelligence being spread on social media, saying that social media companies are Ensure that the material is not dispersed across platforms that have a clear role in enforcing policies to block it.
But Reynolds said it's not just situations involving non-consensual sexual images that can cause harm, especially for people whose names and faces are brands.
CBC News requested interviews with Berg and Mercer and asked whether they had taken any action against the false advertisements that plagiarized their likenesses. Mercer declined to be interviewed for this article. Berg's spokesperson forwarded the request to CTV's parent company, Bell Media, which declined.
13:52Will Taylor Swift's AI deepfake finally force governments to take action?
Last week, an explicit image of an AI-generated portrait of Taylor Swift was shared on X, formerly known as Twitter, without her consent. These photos racked up millions of views before being removed. Reporters Sam Cole and Melissa Heikkila, who have been tracking the rise of deepfakes for years, discuss why this article struck a nerve in Hollywood and Washington.
new legal situation
Pablo Tseng, an intellectual property lawyer with Vancouver-based Macmillan LLP, says it doesn't matter whether someone is famous or not, as long as their image is used in a way they haven't consented to.
“You have control over how you present yourself,” said Tseng, a partner at Macmillan Law Firm. “The law will continue to view this as a fraud committed against you. The question, of course, is do you think this is worth pursuing in court?”
Although Canada has not followed the United States in enacting new deepfake laws, there are several existing torts, and these laws primarily require judges to compensate those harmed by wrongdoing. It has been enacted with the purpose of providing the following benefits, and may be applied in lawsuits involving AI. Tseng said he generated a deepfake.
He said the tort of misappropriation of personality could apply because someone's image is often digitally manipulated or grafted onto another image.
The tort of false light, which relates to publicly misrepresenting a person, is a recent option under U.S. law and was recognized for the first time in the Canadian High Court. in 2019. But so far, it's only recognized in two provinces (the other being British Columbia).
Andrew Chan details the consequences of faking high-profile photos after several recent AI images have gone viral, including the Pope in a fur coat and former President Donald Trump being arrested. do.
play the long game
Reynolds said anyone who wants to take any legal action regarding the production and distribution of deepfakes will need to litigate for a long time. As with any litigation, the court process can be time-consuming and expensive.
But this fight could pay off.
Reynolds pointed to the recent Class action lawsuit against meth From 2011 to 2014, issues surrounding “Sponsored Stories” ads on Facebook that used users' names and profile photos to promote products without their consent.
Last month, Meta offered a $51 million settlement to its Canadian users. Lawyers estimate that 4.3 million people whose real names or photos were used in sponsored articles could qualify.
“While not particularly readily available for individuals, filing a class action lawsuit can be more cost-effective,” Reynolds said.
But seeking justice and reparations also requires knowing who is responsible for these deepfake scams. Lee, of the University of Ottawa, said what is already difficult will become nearly impossible with further advances in generative AI technology.
Much of the research published on artificial intelligence includes openly accessible source code, which means anyone with the know-how can write their own programs without traceable markers, she explained.
AI-generated fake videos are used for scams and internet gags, but what if they're created to disrupt elections? CBC's Catherine Tunney explains how this technology could be weaponized We examine whether Canada is ready for deepfake elections.
