How judges, not politicians, could dictate America’s AI rules
MIT Technology ReviewJul 17, 2023
A handful of major publishers are reportedly close to joining together to file a lawsuit against artificial intelligence (AI) platforms, as well as request legislative action to regulate the use of the companies’ data for training of AI models.
The New York Times and News Corp., owner of the Wall Street Journal, are among the larger entities in the group. Joey Levin, CEO of IAC, another company in the group, warned about how an AI takeover of the media "could be more profound than" "whether AI is going take over the world to eliminate humans and all that stuff."
With legislative action unlikely to come out of Congress, the best ways to act against AI platforms that may be violating the law are lawsuits. With even the threat of lawsuits hanging over them, AI companies should be motivated to build and train their AI in a way that’s fairer and more equitable, and may even get content creators compensated.
With so much data being used to train these AI products, it will be difficult to identify what content is being used and who originally produced them. Winning a lawsuit — which would be costly to both the winner and the loser — could prove difficult. In addition, punishing AIs for using others’ work for learning could lead to similar claims between humans and those who train by watching them.
There's a 97% chance that AI will be able to watch a movie and tell you accurately what is going on before 2030, according to the Metaculus prediction community.