Privacy watchdog says ChatGPT still not meeting data accuracy standards

A task force from the EU’s privacy watchdog said efforts to reduce OpenAI’s ChatGPT chatbot from generating less factually incorrect output are not enough to ensure full compliance with EU data rules.

“While the measures taken to comply with the transparency principle are beneficial to avoid misinterpretation of ChatGPT’s output, they are insufficient to comply with the data accuracy principle,” the task force said in a report released on its website on Friday. Are not.” ,

The body uniting Europe’s national privacy watchdogs set up a task force on ChatGPIT last year after national regulators, led by Italy’s authority, raised concerns about the widely used artificial intelligence service.

OpenAI did not immediately respond to a Reuters request for comment.

The report said that various investigations initiated by national privacy watchdogs in some member states are still ongoing, so it is not yet possible to provide full details of the results. The findings were to be understood as a ‘common denominator’ between national authorities.

Data accuracy is one of the guiding principles of EU data protection rules.

“Indeed, due to the probabilistic nature of the system, current training approaches lead to a model that may produce biased or even fabricated outputs,” the report said.

“Further, the outputs provided by ChatGPT may be considered by end users to be factually accurate, including information relating to individuals, regardless of their actual accuracy.”

© Thomson Reuters 2024

(This story has not been edited by NDTV staff and is auto-generated from a syndicated feed.)