The European Data Protection Board (EDPB) has released its first progress report on OpenAI’s “GPT taskforce,” and the results are unfavorable for ChatGPT. OpenAI’s efforts to align its flagship AI model, ChatGPT, with European Union regulations, such as the General Data Protection Regulation (GDPR), have been acknowledged by the EDPB but ultimately deemed inadequate.
According to the EDPB document, this conclusion follows OpenAI’s struggles in 2024 with temporary halt orders from various European member states. In January, Italy’s data protection agency discovered that ChatGPT and OpenAI were still violating Italian and EU data privacy laws, despite being previously warned and subsequently banned in March 2023.
The EDPB report states that OpenAI has not made sufficient progress in bringing ChatGPT in line with EU regulations. The primary concern is that ChatGPT frequently produces inaccurate information. The EDPB explains, “Due to the system’s probabilistic nature, the current training approach results in a model that may generate biased or fabricated outputs.”
The report also expresses the EDPB’s concern that end users may accept the outputs provided by ChatGPT as factually accurate, regardless of their accuracy. It remains unclear how OpenAI can ensure ChatGPT’s compliance. For instance, the GPT-4 model consists of billions of data points and approximately a trillion parameters. It would be impractical for humans to meticulously review the dataset to verify its accuracy to a degree that satisfies GDPR standards.
Regrettably for OpenAI, the EDPB explicitly states that “technical impossibility cannot be invoked to justify non-compliance with these requirements.”
Relatedly, OpenAI has suspended ChatGPT voice, which was accused of mimicking Scarlett Johansson.