Programming

OpenAI accuses the New York Times of hacking ChatGPT

OpenAI accuses the New York Times of hacking ChatGPT

What in between OpenAI and the New York Times that there is a less than idyllic relationship has been known for some time.

In fact, just at the end of last year, the lawsuit brought by the newspaper against the Sam Altman. Despite this, few would have predicted what happened just a few hours ago.

OpenAI, in fact, recently publicly stated that the NYT would have paid someone to hack its company’s products, including ChatGPT. According to what OpenAI claimed, the newspaper would have carried out tens of thousands of interactions with the chatbotlooking for anomalous results or any useful bugs precisely in the context of the aforementioned legal procedure.

On the other hand, the company is the spearhead of the entireArtificial intelligence insists that ChatGPT does not compete with traditional journalism platforms. Despite this, OpenAI admitted that NYT articles were also used to train its AI models as early as 2020.

How would the New York Times actually act to “force” ChatGPT?

The response from the prestigious newspaper did not take long to arrive. In fact, one of the main legal consultants of the newspaper expressed himself in this regard, namely Ian Crosby. According to the expert “What OpenAI incorrectly defines as ‘hacking’ is simply using OpenAI products to look for evidence that they have stolen and reproduced the Times’ copyrighted works“.

Crosby himself then added how “Creating new content is no excuse for violating copyright laws, and that’s exactly what OpenAI has done on an unprecedented scale“.

According to what OpenAI claimed, the NYT would have asked ChatGPT repeated questions, asking the chatbot to expand its answers until it did not obtain content potentially attributable to newspaper articles or cause anomalous behavior, such as regurgitation of training data o hallucinations of AI. OpenAI’s accusation is based precisely on the potential exploitation of these “bugs”.

Beyond these skirmishes, it seems clear that the trial between the two parties could mark a turning point in the relationship between more classic journalism and emerging technologies.

Source: arstechnica.com

Leave a Reply

Your email address will not be published. Required fields are marked *