OpenAI Seeks International Governance of Most Powerful AI Systems

OpenAI wants the development of artificial intelligence (AI) to include widespread input and regulation.

Greg Brockman, the president of the ChatGPT creator, said Monday (May 22) that OpenAI is considering a Wikipedia-like model that would enable the sharing of diverse views and the arrival at a consensus on decisions around AI, Reuters reported Monday.

“We’re not just sitting in Silicon Valley thinking we can write these rules for everyone,” Brockman said at an AI Forward event in San Francisco hosted by Goldman Sachs Group and SV Angel, according to the report. “We’re starting to think about democratic decision-making.”

Brockman’s comments were made on the same day OpenAI published a blog post authored by Brockman, CEO Sam Altman and Chief Scientist Ilya Sutskever that said it’s time to start thinking about the governance of future AI systems.

In their post, the three authors suggested that the leading development efforts in AI be coordinated to limit the rate of growth per year in AI capability, that an international authority be formed to monitor AI development efforts and restrict those above a certain capability, and that technical capability be developed to make superintelligence safe.

The authors added that the regulations should not be applied to the development of AI models below a certain threshold of capability.

“But the governance of the most powerful systems, as well as decisions regarding their deployment, must have strong public oversight,” Altman, Brockman and Sutskever said in the blog post. “We believe people around the world should democratically decide on the bounds and defaults for AI systems. We don’t yet know how to design such a mechanism, but we plan to experiment with its development.”

Two days earlier, on Saturday (May 20), the leaders of the G7 nations said in a bulletin that they want to hold talks on reaching “common vision and goal of trustworthy AI.”

The seven countries — the United States, the United Kingdom, Canada, France, Germany, Italy and Japan — plan to convene a summit on AI later this year.

“These discussions could include topics such as governance, safeguard of intellectual property rights including copyrights, promotion of transparency, response to foreign information manipulation, including disinformation, and responsible utilization of these technologies,” the bulletin said.