A coalition of major Canadian legacy media outlets is suing OpenAI, accusing the tech company of using its news articles to train ChatGPT without the collective’s permission.
The Canadian corporate press outlets include the CBC, Post Media, The Globe and Mail, The Canadian Press, Toronto Star and Toronto Star. They have submitted a statement of claim against OpenAI to the Superior Court of Justice on Friday.
According to the claim, the legacy media coalition alleged that OpenAI is “severally liable” for Copyright Act infringements by using the company’s intellectual property, news articles, to train its artificial intelligence software ChatGPT.
They alleged that the tech company and its conglomerates have “unjustly enriched themselves at the expense of the News Media Companies” and have breached the company’s terms of service by copying its articles for the alleged use.
The group is demanding OpenAI pay $20,000 per article it used without consent, with more for damages to be determined by the court, if the suit is successful.
The suit also seeks a permanent injunction to prevent OpenAI and anyone affiliated with the company from directly or indirectly infringing on those copyrights again.
“The Defendants have engaged in ongoing, deliberate, and unauthorized misappropriation of the Plaintiffs’ valuable news media works. The Plaintiffs bring this action to prevent and seek recompense for these unlawful activities,” it said.
It alleges that as ChatGPT works off of pattern recognition by analyzing enormous quantities of text data, the training of that software “deliberately scrapes” content from the legacy media, allegedly using the proprietary content without consent to develop its models.
This is the first time the five Canadian media organizations have banded together to sue the tech company.
E-Law professor, Michael Geist told True North he will be keeping an eye on the “major case” as it unfolds.
He said the lawsuit is the second copyright AI case launched in a month in Canada.
The Canadian Legal Information Institute, or CanLII, has recently filed a suit against Caseway AI for allegedly infringing on its copyrights.
“Similar cases in the US haven’t gotten much traction yet but are ongoing,” he said.
OpenAI is currently involved in several lawsuits in the U.S. over alleged copyright infringements that involved training its AI models.
The Authors Guild and 17 authors, including George R. R. Martin, filed a class action against the company in September of last year.
In December 2023, the New York Times launched litigation, and last April, eight separate US outlets owned by Tribune Publishing filed separate lawsuits against the company, citing again that their intellectual property rights were infringed to train the AI software.
“Indeed, OpenAI recently succeeded in defeating a case against a couple of news organizations in a NY case,” Geist said.
This month, a judge dismissed a copyright infringement lawsuit against OpenAI by two news outlets, Raw Story and Alternet. The judge ruled that the two outlets could not demonstrate sufficient harm to continue the case. However, the two can refile their complaint as the decision was made without prejudice.
The Liberal Government made similar claims, using the Online News Act, against Google for using the intellectual property of Canadian media organizations to populate its search engine. Ultimately, Google agreed to allow various news outlets to use its content.
Geist speculates that the litigation against OpenAI could serve a similar purpose and result in a settlement to force the company into a licensing agreement with the aggrieved outlets.