The Group of Seven (G7) industrial nations is poised to reach an agreement on an AI code of conduct for developers on October 30, according to Reuters. This set of 11 guidelines seeks to foster "safe, reliable, and trustworthy AI globally" while addressing and mitigating the associated risks, aiming to harness AI's advantages. The plan, which was initially drafted by G7 leaders in September, offers voluntary action recommendations for organizations involved in the development of state-of-the-art AI systems, including advanced base models and generative AI systems.
Among its recommendations, the code of conduct encourages companies to publish comprehensive reports about the capabilities, limitations, applications, and potential misuses of the AI systems they create. Furthermore, it advocates for robust security controls on these systems. G7 member nations include Canada, France, Germany, Italy, Japan, the United Kingdom, the United States, and the European Union. This year's G7 meeting is taking place in Hiroshima, Japan, with digital and technology ministers from participating countries convening from April 29 to 30.
Topics on the conference agenda encompass emerging technologies, digital infrastructure, and artificial intelligence. Additionally, a dedicated agenda item focuses on responsible AI and global AI governance. The G7's AI code of conduct emerges in a broader global context where governments are responding to the emergence of AI, aiming to harness its benefits while addressing concerns. Notably, the European Union was among the first to establish guidelines, as seen in the EU Artificial Intelligence Act, whose initial draft was adopted in June.
On October 26, the United Nations formed a 39-member advisory committee to tackle issues associated with the worldwide regulation of artificial intelligence. China has also introduced its own AI regulations, effective as of August. Within the AI industry, OpenAI, known for its AI chatbot ChatGPT, announced plans to form a "readiness" team tasked with assessing various AI-related risks.
