Japan’s privacy watchdog directed OpenAI to refrain from gathering sensitive personal data

Japan’s privacy watchdog raises concerns over OpenAI’s data collection practices. The watchdog has advised OpenAI to refrain from collecting sensitive data without individuals’ consent. Specifically, the Personal Information Protection Commission has cautioned OpenAI to minimise the amount of sensitive data it gathers for machine learning purposes. The commission has directed the company to refrain from gathering data related to individuals’ thoughts, beliefs, medical history, and criminal records without their explicit consent. OpenAI has been advised to take measures to ensure that the data collected for machine learning purposes does not contain sensitive content. Furthermore, any such sensitive information discovered during data processing should be promptly deleted. On the other hand, the watchdog acknowledges the need to balance privacy concerns with the potential benefits of generative AI, such as driving innovation and addressing challenges like climate change.

This warning reflects the global effort among regulators to establish regulations governing the use of generative AI. Within the EU, this technology has faced temporary bans in Italy and is currently under investigation in other European countries, including Spain and Germany, among others. While recently, in a similar development, Canadian privacy regulators launched a joint investigation into OpenAI’s data collection.

What are the potential consequences if OpenAI continues to collect sensitive data without permission?
Failure by OpenAI to obtain permission to collect sensitive data may result in regulatory consequences, including substantial fines, mandated data deletion, or even a potential ban on its operations.

What is the Japanese vision for AI?

Japan has recently unveiled the draft of its Integrated Innovation Strategy for 2023, which focuses on promoting and enhancing domestic capabilities in generative AI development. The strategy recognises the potential of generative AI in boosting productivity and addressing societal challenges. However, it also highlights the importance of addressing the associated risks. The draft aims to tackle risks such as copyright infringement, exposure of confidential information, dissemination of false information, and vulnerability to cyberattacks. International discussions will play a crucial role in examining and addressing these AI-related risks.