The European Data Protection Board (EDPB) has released its initial progress report on OpenAI’s “GPT taskforce,” and the findings are unfavorable for ChatGPT.
OpenAI’s attempts to align its flagship AI model, ChatGPT, with European Union regulations, including the comprehensive General Data Protection Regulation (GDPR), have been acknowledged by the EDPB. However, they have ultimately been deemed inadequate.
According to the EDPB document:
These findings come in light of OpenAI facing temporary stop orders from various European member states throughout 2024.
As reported by Cointelegraph in January, Italy’s data protection agency discovered that ChatGPT and OpenAI were still violating Italian and EU data privacy laws, despite being warned and subsequently banned in March 2023.
The EDPB’s report states that OpenAI has not done enough in the intervening period to ensure ChatGPT’s compliance with EU laws.
The primary concern seems to be that ChatGPT often produces inaccurate information.
“As a matter of fact,” states the EDPB, “due to the probabilistic nature of the system, the current training approach leads to a model which may also produce biased or fabricated outputs.”
The report also expresses the EDPB’s worry that “end users are likely to perceive the outputs provided by ChatGPT as factually accurate, regardless of their accuracy.”
At this time, it remains unclear how OpenAI can bring ChatGPT into compliance. The GPT-4 model, for instance, consists of billions of data points and around a trillion parameters. It would be impractical for humans to meticulously scrutinize the dataset to ensure its accuracy to a degree that satisfies GDPR standards.
Unfortunately for OpenAI, the EDPB explicitly states that “claims of technical impossibility cannot be used to justify non-compliance with these requirements.”
Related:
OpenAI suspends ChatGPT voice accused of mimicking Scarlett Johansson
EU finds ChatGPT's compliance with regulations compromised due to excessive dissemination of incorrect data