tech
OpenAI may have made a fatal misstep in copyright fight with news orgs
OpenAI may be sanctioned for hiding, deleting ChatGPT logs in NYT copyright fight.
TL;DR
- News organizations, led by The New York Times, are seeking "serious sanctions" against OpenAI.
- They accuse OpenAI of repeatedly lying and concealing evidence regarding users circumventing paywalls by prompting ChatGPT to regurgitate copyrighted articles.
- An OpenAI privacy engineer's deposition allegedly revealed that the company misled the court for two years about the cost and burden of searching ChatGPT logs.
- OpenAI is accused of pretending it lacked the technical ability to search logs when it had already conducted such searches.
- News plaintiffs argue OpenAI's actions withheld relevant evidence, prolonged discovery, inflated expenses, and burdened the court.
- OpenAI claims the sanctions motion is a tactic by the NYT to access more logs and infringe user privacy, suggesting the NYT's case is weakening.
- NYT disputes their case is weakening, stating it was streamlined and strengthened by adding claims against Microsoft.
- Allegedly, OpenAI had two large samples of logs (10 million and 78 million) that had been de-identified and could have been provided to plaintiffs earlier.
- OpenAI also allegedly searched these samples for NYT content as part of research into blocking copyrighted content regurgitation.
- Instead of transparency, OpenAI allegedly forced plaintiffs to search a heavily redacted sample of 20 million logs, which was further skewed by AI redactions.
- News plaintiffs accuse OpenAI of deleting or compressing billions of logs that should have been preserved, despite a court order.
- The news organizations want OpenAI prohibited from using the 20 million sample, a finding that withheld logs contained regurgitated content, and jury instructions about deleted logs.
- They argue that lesser sanctions would not be effective and that OpenAI's misconduct was knowing and intentional.