The Case for an IPCC Clone to Help Govern AI

The Case for an IPCC Clone to Help Govern AI

Does the world need a supranational organization to track developments in artificial intelligence (AI) and advise governments on responding to the technology’s rapid advance?

Experts, including Eric Schmidt, former chief executive of Google, and Ian Bremmer, president of the Eurasia Group, call for such an organization to be created. They propose a body modeled on the Intergovernmental Panel on Climate Change (IPCC), perhaps called the International Panel on AI Safety (IPAIS).

It is an intriguing idea, but would such a body provide the support governments need to keep pace with the breakneck speed at which AI is progressing?

There is a need for a global mechanism that monitors AI developments and offers concise, layperson-level advice on how to govern the adoption of this transformative technology. AI can be technically challenging, and its applications are multiplying so fast that policymakers struggle to keep abreast of them.

Also, while national initiatives to govern AI may be laudable, they are inadequate. AI’s impact is worldwide, and policies to manage it must reflect its globe-spanning reach. A patchwork of national initiatives to govern AI could stifle its progress as could over-regulation generally. A single, authoritative source of information on the technology could steer regulators away from unrealistic and/or damaging restrictions on the technology.

Another issue that an international advisory group might help address is the secretive nature of AI’s development. There is concern that a few tech giants control the future of a technology that has societal impacts. A noncommercial, supranational body of experts could represent the wider community’s interests and strive to ensure that AI’s progress aligns with the common good.

On the surface, it seems like an IPPC-like body might be the right vehicle to meet these demands.

Created in 1988, the IPPC is a United Nations body with 195 members. The organization is divided into three Working Groups and a Task Force. Thousands of people worldwide contribute to its reports, which assess thousands of scientific papers published annually. These reports summarize what is known about the drivers of climate change, its impacts and future risks, and how adaptation and mitigation can reduce those risks. Governments across the globe use this information to guide their policymaking.

The organization is also a global platform for discussion. IPCC member government representatives meet at least yearly in plenary sessions. Government officials from member countries, experts, and research institutions attend the sessions.

Another important function the panel fulfills is raising awareness of climate change issues. Its reports attract substantial media coverage and discussion.

However, the IPCC also has some serious downsides. Many people widely accuse it of devolving into a talking shop that fails to achieve meaningful change. Governments often ignore or delay its recommendations or use them in their greenwashing campaigns. These shortcomings are symptomatic of many UN programs. Nation-states that supposedly support these programs are the ones that hold the power, not organizations like the IPCC.

On balance, an AI equivalent to the IPCC model merits serious consideration. At the very least, it could focus attention on AI issues that urgently need to be tackled. Given the pace at which AI is moving, the body would have to meet much more frequently than the IPCC does and be more agile than the climate change body. Perhaps other types of international organizations, such as Interpol, that report to nation-states might provide a more realistic model.

Either way, we need to act soon. The longer we delay, the more difficult it will be to govern a technology that is changing the world.

The idea of an International Panel on AI Safety resonates deeply with our mission. As we navigate the complexities of AI in plagiarism detection and beyond, there is a strong need for something to give companies guidance on what is right, wrong, and how to use these technologies both ethically and responsibly.

Like
Reply
Alan Ross

Chairman of What Can You Do ( Earth) C.I.C..,President of International Association of AI -agent Developers and CEO of Global-AI Solutions

4mo

AI safety concerns of the more thoughful leaders of the AI community are well placed and there is clearly a need and a growing momentum to establish an IPCC type body to provide some oversight of AI developments to make sure it benefits all humanity . Sadly the rapid pace of development and the intense competion between big Tech companies and betwen countries will make this very difficult , Basic rules need to be established to outlaw misuse of ai such as false videos etc

Like
Reply
Lawal Audi

Independent Construction Professional

4mo

Useful tips

Like
Reply
Dr. Razon Chandra Saha

Freight Transport Specialist

4mo

YossiSir @ Meaningful and we need to take utmost care about AI

Anil Bhandari

Independent Consultant, Infrastructure.

4mo

A great proposal that makes a lot of sense, Yossi.

To view or add a comment, sign in

More articles by Yossi Sheffi

Insights from the community

Others also viewed

Explore topics