G7 nations agree on 'risk-based' artificial intelligence regulation

Group of Seven advanced nations should adopt “risk-based” regulation on artificial intelligence, their digital ministers agreed on Sunday, as European lawmakers hurry to introduce an AI Act to enforce rules on emerging tools such as ChatGPT.

But such regulation should also “preserve an open and enabling environment” for the development of AI technologies and be based on democratic values, G7 ministers said in a joint statement issued at the end of a two-day meeting in Japan.

While the ministers recognised that “policy instruments to achieve the common vision and goal of trustworthy AI may vary across G7 members”, the agreement sets a landmark for how major countries govern AI amid privacy concerns and security risks.

“The conclusions of this G7 meeting show that we are definitely not alone in this,” European Commission Executive Vice President Margrethe Vestager told Reuters ahead of the agreement.

Governments have especially paid attention to the popularity of generative AI tools such as ChatGPT, a chatbot developed by Microsoft-backed OpenAI that has become the fastest-growing application in history since its November launch.

“We plan to convene future G7 discussions on generative AI which could include topics such as governance, how to safeguard intellectual property rights including copyright, promote transparency, address disinformation” including information manipulation by foreign forces, the ministerial statement said.

Italy, a G7 member, took ChatGPT offline last month to investigate its potential breach of personal data rules. While Italy lifted the ban on Friday, the move has inspired fellow European privacy regulators to launch probes.

EU lawmakers on Thursday reached a preliminary agreement on a new draft of its upcoming AI Act, including copyright protection measures for generative AI, following a call for world leaders to convene a summit to control such technology. Vestager, EU’s tech regulation chief, said the bloc “will have the political agreement this year” on the AI legislation, such as labelling obligations for AI-generated images or music, to address copyright and educational risks.