"Why do AI Technology Executives Hold Closed-door Meetings and What Discussions are Taking Place? Are They Afraid of Regulatory Legislation?"

The latest AI Insight Forum, a gathering of elites in the technology industry, was held on Capitol Hill, resembling a gathering of billionaires. This closed-door meeting brought together all the tech giants, with a combined net worth of approximately $550 billion, to discuss the future of artificial intelligence, with a particular focus on the controversial topic of "AI regulation." Unlike previous AI Senate hearings, which included testimonies from individuals such as Sam Altman from OpenAI and Dario Amodei, CEO of Anthropic, this recent meeting assembled prominent figures like Elon Musk and Bill Gates, and was not open to the public or the media. This raises the question of what important or controversial discussions took place during this closed-door meeting?

What's next?

Senator Chuck Schumer of New York was one of the 60 senators who participated in the meeting, describing it as a "highly productive inaugural AI Insight Summit" and stating that the future tasks are challenging. Congress seeks to pass a bipartisan AI legislation next year aimed at addressing AI-related risks.

The tech giants who attended the event highly praised the open discussion on AI regulation that the meeting promoted. Figures such as Sam Altman, Mark Zuckerberg, CEO of Meta, Jensen Huang, CEO of NVIDIA, Satya Nadella, CEO of Microsoft, Sundar Pichhai, CEO of Alphabet, Arvind Krishna, CEO of IBM, Alex Karp, CEO of Palantir, former Google CEO Eric Schmidt, Clement Delangue, CEO of HuggingFace, and Tristan Harris, co-founder and executive director of the Center for Humane Technology, were among the attendees. Elon Musk described the meeting as an event serving humanity and emphasized the need for a "referee" to ensure companies take safe actions and protect the public interest.

Zuckerberg also expressed his views after the meeting, stating that Congress should collaborate with artificial intelligence to promote innovation and establish protective measures, believing that cooperation between the government and large tech companies would be beneficial in addressing such issues.

In previous discussions on artificial intelligence, leaders and senators seemed to agree on regulations that could mitigate AI risks. However, the question of how to implement these regulations remains unresolved.

Closed-door permit system

Senator Schumer stated that the closed-door format facilitated open discussions among participants, abandoning the time and format restrictions imposed by general public hearings, and mentioned that some future discussions will be open to the public.

Pedro Domingos, a computer science professor at the University of Washington, hinted at the possibility of introducing a licensing system for running large language models.

Congress is considering the idea of requiring a government license to learn and run LLMs, which is absurd to the extreme. It could happen.
- Pedro Domingos (@pmddomingos) September 13, 2023

The permit system could have two possible outcomes. Through the permit system, large tech companies would proceed cautiously with AI development, potentially becoming catalysts for responsible AI. However, if obtaining a permit becomes a mandatory requirement, AI development for small tech companies could be hindered, causing them to fall behind in competition. Therefore, the complexity of AI regulations cannot be generalized for all companies. Interestingly, Sam Altman previously mentioned that small companies should not be regulated.

Looking back a few months ago, when some leaders signed a petition to slow down the development of AI, urging OpenAI not to train advanced models like GPT-5, it is the large companies that would be affected if any form of regulation is enacted. Hence, the tech giants held this meeting.

Proactive prevention

In the past, the US government's attitude towards regulating new technologies has always been unclear. For example, autonomous vehicles are not fully regulated in the United States. Despite accidents caused by these vehicles, there is still no clear legislation regarding autonomous driving. Efforts by Congress to enact legislation for autonomous vehicles have faced years of delays. Similarly, regulations for cryptocurrencies only began after several years in the market, during which fraud and misallocation of funds were issues, indicating a preventive approach by the US government towards AI regulation.

The government is attempting to avoid the precedent of certain technologies/products causing negative impacts when unregulated, taking a proactive approach to AI regulation.

The government hopes to control the development of technology before the situation worsens. Additionally, due to numerous predictions by AI researchers and experts about the potential doomsday scenarios, this issue has received significant attention. Interestingly, in an interview in 2016, Musk told Altman that it is necessary to democratize AI technology and that AI could be used for negative purposes.

Based on the discussions initiated by the Senate four months ago and initiatives such as OpenAI's aim to democratize AI, there have been no substantial results yet. In this process, not only the United States but also companies worldwide continue to release advanced models. It seems that if any form of regulation is necessary, it is best to implement it as soon as possible.