OpenAI says allegations ‘blatantly false,’ asserts fair use
More than a dozen news organizations, including The New York Times, the New York Daily News and The Intercept, asked a federal court on Thursday to impose sanctions on OpenAI, accusing the company of lying about its ability to search its training data and of deleting evidence in lawsuits over copyright infringement. The news organizations have sued OpenAI, alleging the company violated copyright law by using their articles to train its AI models without permission or payment.
In their court filing, the news organizations said OpenAI previously told the court it could not search its training datasets and ChatGPT output logs for copyrighted material. But an OpenAI employee said during a deposition earlier this year that the data could be accessed, according to the filing. “The evidence is in OpenAI’s training data sets and ChatGPT output logs,” the organizations wrote. “But instead of just producing that evidence at the start of the case and focusing on the merits of its fair use defense, OpenAI chose obstruction.”
The filing also alleges that OpenAI deleted data logs in violation of a court order requiring the preservation of relevant evidence. An attorney for the news organizations said in a statement that OpenAI had claimed searching its ChatGPT outputs was “infeasible, burdensome and invasive of users’ privacy” but then lied about having already done such searches.
OpenAI responded by calling the allegations “blatantly false.” The company said its use of copyrighted content to train its models falls under “the long-established principles of fair use,” according to a statement cited in the filing.