First reported by The Guardian, a number of major Canadian news and media companies have banded together to sue OpenAI over its use of their articles to train large language models.
In a statement about the lawsuit, News Media Canada president Paul Deegan argued that "these artificial intelligence companies cannibalize proprietary content and are free-riding on the backs of news publishers who invest real money to employ real journalists who produce real stories for real people.
"They are strip-mining journalism while substantially, unjustly, and unlawfully enriching themselves to the detriment of publishers."
The suit was filed on Friday, and calls for a share of any profits OpenAI made from the use of articles from these companies, an injunction on OpenAI's continued use of any content from them, and damages of up to $20,000 per article used by OpenAI to train its LLMs. Given the dragnet nature of AI model training and the sheer number of individual articles likely in question, OpenAI could be liable for catastrophic damages if the court rules in the media companies' favor. The companies behind the lawsuit include:
I'm not usually one to lose any sleep over the protection of corporate copyright, but it's starting to become clear that copyright law may prove an effective defense against AI companies swallowing up the internet whole and spitting it back out to us in diminished form. OpenAI is currently also fending off copyright lawsuits from the New York Times and a class action of authors including George R.R. Martin, while Elon Musk has also sued OpenAI in a bit of palace intrigue between him and other co-founders of the supposed non-profit.
