Nobody Knows How to Regulate ChatGPT

The EU is as stumped as everyone else on how to regulate generative AI. EU regulators were in the middle of drafting the world’s first binding AI rulebook, the AI Act when ChatGPT was released. “My God, our draft is already out of date,” said a European Parliament member who was handling the AI legislative proposal. The Act was to have entrusted the European Commission with establishing cybersecurity, transparency and risk-management requirements for general-purpose AIs. Now everything has changed. The EU Council wrote an open letter to the Commission’s President and U.S. President Joe Biden, calling on them to organize a global AI summit. 

The purpose of the summit would be to discuss the governing principles that would control the development and deployment of AI models, making sure they are human-centric, safe and trustworthy. Canada, France, Italy and Spain each have launched investigations into OpenAI’s ChatGPT because of data privacy issues. Italy has temporarily blocked access to the AI chatbot amid concerns about protecting personal data and minors. And the United States will be holding congressional hearings on forcing assessment of AI products before companies release them. At this point, the  Commission’s proposed draft rules for an AI Act have grown from 52 to 108 pages. A parliamentary committee “hopes to at least reach a common position with all parties” by April 26, 2023.