OpenAI 'accidentally' deletes ChatGPT evidence, adding to tension in lawsuit with New York Times

OpenAI's Questionable Move in Copyright Lawsuit

According to TechCrunch , the copyright lawsuit between the New York Times and OpenAI is taking a new turn when lawyers for the New York Times (NYT) accused OpenAI of deleting ChatGPT training data related to suspected copyright infringement content. This move comes after experts from the NYT and Daily News spent more than 150 hours searching for evidence in the training data set provided by OpenAI.

According to a letter sent to the Southern District of New York on November 14, data on one of the two virtual machines that OpenAI provided to the legal team was deleted by the company’s engineers. Although OpenAI said this was an unintentional technical error, the New York Times and other plaintiffs said this severely compromised their efforts to find evidence.

OpenAI sued by New York Times and several other publishers for alleged copyright infringement by using content to train ChatGPT without permission

PHOTO: SCREENSHOT THE ECONOMIC TIMES

OpenAI had previously agreed to let the plaintiffs inspect the ChatGPT training data in a secure environment, a major step in the publishers’ efforts to prove that their content was improperly used to train AI models. But the deletion of the data meant that the lawyers lost access to important information.

The deleted data has been recovered, but in a format that is legally unusable, making it difficult for the New York Times to present evidence in court. The move raises questions about whether OpenAI has the capacity and commitment to ensure the integrity of the data, especially in the context of rising copyright infringement lawsuits.

Reaction from stakeholders

The New York Times said the deletion of the data severely disrupted the review process and could be seen as a lack of transparency. OpenAI, meanwhile, insisted that it was a technical glitch and not related to an attempt to conceal evidence. Still, the incident highlights the increasingly complex legal challenges in holding AI companies accountable for using copyrighted data to develop artificial intelligence models.

The lawsuit could have major implications for how AI companies manage data and comply with regulatory requirements in the future. Now, news publishers are considering further steps to protect their interests, while OpenAI faces pressure to improve its data storage and vetting processes to avoid similar mishaps.

Source: https://thanhnien.vn/openai-vo-tinh-xoa-chung-cu-chatgpt-khien-vu-kien-voi-new-york-times-them-cang-thang-185241122213504967.htm