ChatGPT faces Austria complaint over ‘uncorrectable errors’

A Vienna-based privateness marketing campaign group mentioned Monday it will file a criticism in opposition to ChatGPT in Austria, claiming the “hallucinating” flagship AI device has invented incorrect solutions that creator OpenAI can not appropriate.

NOYB (“None of Your Business”) mentioned there was no method to assure this system offered accurate information. “ChatGPT keeps hallucinating—and not even OpenAI can stop it,” the group mentioned in a press release.

The corporate has brazenly acknowledged it can not appropriate inaccurate info produced by its generative AI device and has failed to elucidate the place the information comes from and what ChatGPT shops about people, mentioned the group.

Such errors are unacceptable for details about people as a result of EU legislation stipulates that personal data should be correct, NOYB argued.

“If a system cannot produce accurate and transparent results, it cannot be used to generate data about individuals,” mentioned Maartje de Graaf, data-protection lawyer at NOYB.

“The know-how has to comply with the legal requirements, not the opposite approach round.”

ChatGPT “repeatedly provided incorrect information” concerning the start date of NOYB founder Max Schrems “instead of telling users that it doesn’t have the necessary data”, mentioned the group.

OpenAI refused Schrems’s request to rectify or erase the information regardless of it being incorrect, saying it was not possible, NOYB added.

It additionally “failed to adequately respond” to his request to entry his private knowledge, once more in violation of EU legislation, mentioned NOYB, and the agency “seems to not even pretend that it can comply”.

OpenAI mentioned it was “committed to protecting data privacy” in response to an AFP request for remark.

“We want our AI models to learn about the world, not individuals; we do not actively seek personal information to train our models, and we do not use publicly available information on the Internet to profile, advertise to, or target people, or to sell their data,” mentioned an OpenAI spokesperson.

NOYB, which has emerged as a fierce critic of tech giants since its creation in 2018, mentioned it was asking Austria’s knowledge safety authority to research and advantageous OpenAI to deliver it in step with EU legislation.

Bursting onto the scene in November 2022, ChatGPT sparked a frenzy amongst tech customers dazzled by its capability to reel off dissertations, poems or translations in mere seconds.

However criticism of the know-how has prompted legal action in some nations.

Italy quickly blocked this system in March 2023, whereas France’s regulatory authority started an investigation after a collection of complaints.

A European working group has additionally been set as much as enhance coordination, though NOYB mentioned it was skeptical concerning the authorities’ efforts to manage AI.

© 2024 AFP