The UK calls on countries to label AI as capable of causing 'catastrophic harm'.

Oct 19
At an AI safety meeting being held in the UK next month, Prime Minister Rishi Sunak is urging nations to declare artificial intelligence as having the potential to cause "catastrophic harm" to create a unifying global response to the quickly developing technology.

According to a draft communiqué distributed to attendees and obtained by Bloomberg, Britain wants the nations to establish a common position stating specific worries about the impact of AI on cybersecurity and biotechnology. By October 25, officials hope to have the final text of the statement ready.
The most hazardous capabilities of these AI models have the potential to cause enormous, potentially catastrophic harm, both intentionally and unintentionally, according to the draft from October 16.

The memo states that world leaders will also demand more openness from businesses creating artificial intelligence technologies. They will concentrate on establishing tools for security testing, unambiguous evaluation metrics, and relevant public sector and research skills.

The edges of general-purpose AI and, in some situations, limited AI with potentially hazardous capabilities provide particular security threats. The most important of these hazards comes from the possibility of intentional abuse or control problems, where AI systems might want to exert more influence and less control over human affairs. The fact that these capabilities are not fully recognized is one of the causes of these issues, the report claims. the project.

Following the summit, the UK wants to organize follow-up meetings to discuss the development of managing AI potential and risks every six months.

In the draft communiqué, it was said that "AI must be designed, developed, deployed, and used in a way that is human-centered, safe, trustworthy, and responsible."

Created with