UN Scientific Panel Warns with Measures for AI Safety

A scientific panel affiliated with the United Nations stated that governments must bring these technologies under control before the potential risks of AI agents are fully understood. The statement was recorded as the UN's first major assessment on AI safety.

Artificial Intelligence on the Global Diplomacy Agenda

This report further solidified artificial intelligence's place on the global diplomacy agenda this week as leaders gathered in New York for the United Nations General Assembly and the US and China held talks on AI.

Last week, UN Secretary-General António Guterres called on governments to cooperate in combating the threats posed by artificial intelligence, stating that the world cannot afford a race to the bottom in AI safety.

First Report of the Independent Scientific Panel

Established last year, the Independent International Scientific Panel on Artificial Intelligence holds the distinction of being the first global scientific body in the field of AI. This report serves as the panel's first thematic briefing.

The report demands much more attention and resources be devoted to managing emerging risks from advanced AI, as well as stronger international coordination on safety and accountability even if individual countries adopt disparate legal approaches.

The Precautionary Principle and Safety Measures

The panel argues that the world does not have to wait for scientists to prove how or why such events happen fully. It is stated that the risk of losing control is precisely the kind of problem the precautionary principle addresses where potential harm could be catastrophic or irreversible.

First appearing in the 1992 UN Rio Declaration on Environment and Development, the precautionary principle argues that scientific uncertainty cannot be an excuse to delay taking action against potentially serious or irreversible harm.

Incidents at Companies

Since the initial report of the cyberattack on Hugging Face, various incidents have been documented at companies such as OpenAI, Anthropic, Google, and Meta. These incidents include cyberattacks targeting real-world targets and swarms of AI agents taking over online message boards.