Artificial intelligence company OpenAI is based in San Francisco and registered under the laws of Delaware.Dado Ruvic/Reuters
A copyright lawsuit filed against OpenAI in Ontario by news media organizations should be heard in the United States instead because the company does not conduct business in the province, lawyers for the artificial-intelligence giant argued at a court hearing Wednesday.
Canada’s major news organizations, including The Globe and Mail and the Canadian Broadcasting Corporation, sued OpenAI last November for allegedly violating copyright law by scraping proprietary news content without consent or payment to train its models, such as those that power ChatGPT.
Lawyers for OpenAI said that the Ontario court does not have jurisdiction because none of the corporate entities named as defendants in the suit conduct business in the province or have offices there. OpenAI’s model training and web-crawling activities occur elsewhere, the lawyers said at the Superior Court of Justice hearing.
“The conduct that is barely alleged in the statement of claim overwhelmingly occurs outside of Canada,” said Marc Crandall, a partner at Gowling WLG who represents OpenAI. “Conduct outside of Canada is not within the court’s jurisdiction.”
OpenAI and Meta adjust AI chatbots to better respond to teens in distress
OpenAI is based in San Francisco and registered under the laws of Delaware.
Plaintiffs in the lawsuit also include Postmedia Network Inc., Toronto Star Newspapers Ltd., Metroland Media Group, The Canadian Press and Radio-Canada.
Sana Halwani, a partner at Lenczner Slaght LLP who represents the news companies, said in court that OpenAI’s arguments that the court has no jurisdiction in the matter are “impossible to accept.”
Her colleague, Monique Jilesen, laid out eight ways OpenAI operates in Ontario, including having customers for ChatGPT in the province. “They’ve been very focused on where a server is, and they’re not focused on the big picture of how this business is actually conducted,” she said.
In a written document, the lawyers said OpenAI’s web crawlers have accessed the news companies’ websites, and allegedly broke the terms of service by scraping data for commercial purposes, through servers in Ontario. The company also has a partnership with Microsoft Corp. to sell its products and services in Canada, and its models are “reproduced and hosted” in a Microsoft data centre in Toronto, according to the lawyers.
On that point, OpenAI said that its finished models do not contain copies of any training data.
Judge Jessica Kimmel said she will issue a decision soon.
Lawyers for the news companies wrote in a court document that kicking the matter to the U.S. would amount to “Canada giving up jurisdiction over a large part of its digital economy – a sobering prospect, particularly in the context of news media and its importance to Canadian sovereignty.”
OpenAI called those arguments “hyperbolic” and said “attempts to politicize the motion through appeals to sentiment should be rejected as distractions from the legal issues.”
Bell fighting $400-million lawsuit from film companies alleging it broke copyright law
The lawsuit is one of many filed against AI companies in North America by authors, artists and publishers who argue that training on their intellectual property violates copyright law. Some of the suits also contend AI chatbots can regurgitate copyrighted material verbatim.
AI companies have generally argued that building their models on information culled from the internet without permission is allowed under copyright laws. Canada has a provision called fair dealing, which permits using copyrighted material for research and educational purposes.
But how this provision applies to commercial AI models has proven controversial. The Copyright Act in Canada was conceived well before generative AI took off, and while the federal government completed a consultation last year about updating the legislation, it remains unchanged.
Evan Solomon, the federal minister of AI and digital innovation, previously told The Globe and Mail the government is watching how court cases play out so that legislation is “aligned with market signals to how content creators are compensated.”
Toronto-based AI company Cohere Inc. is facing a lawsuit in the U.S. from news companies, including the Toronto Star, over similar issues. Cohere has denied the allegations and asked the court to dismiss the lawsuit.
link
