British Prime Minister Rishi Sunak has issued a warning against rushing to regulate artificial intelligence (AI) before fully comprehending the technology’s inherent risks. While acknowledging the potential perils associated with AI, including its role in facilitating the development of chemical or biological weapons, its exploitation by terrorist organizations for fear-mongering, or its use by cybercriminals for cyberattacks and fraud, Sunak stressed the need for a nuanced approach to AI regulation. He emphasized that while AI has the capacity to revolutionize life, it is essential to address the potential risks it poses to human existence, akin to pandemics and nuclear warfare. Sunak affirmed that the responsibility for safeguarding against AI’s risks rests primarily with governments rather than the technology companies developing it.
Sunak expressed concerns about AI developers, who may not fully grasp the capabilities their models could ultimately attain, and urged against them being solely responsible for regulating themselves. He asserted that only governments possess the competence to assess national security risks adequately and have the authority and responsibility to ensure the safety of their citizens.
However, Sunak emphasized that the United Kingdom’s approach is not to hastily enact regulations. He questioned how laws could be formulated for a technology that remains not fully understood.
The international community is actively engaged in addressing the challenges associated with AI, especially in light of the emergence of general-purpose AI systems like ChatGPT, which have sparked both excitement and apprehension. Sunak’s forthcoming U.K. AI Safety Summit will concentrate on the risks arising from so-called frontier artificial intelligence, referring to cutting-edge systems with the capacity to perform diverse tasks but potentially harboring undisclosed threats to public safety and security. These systems are built upon large language models, extensively trained on vast repositories of text and data.
One of the primary objectives of the summit is to advocate for the creation of the first-ever international declaration outlining the nature of AI risks, Sunak explained.
In addition to these initiatives, Sunak announced the establishment of an AI Safety Institute, tasked with the examination, evaluation, and testing of new AI varieties. He also proposed the creation of a global expert panel, drawing inspiration from the United Nations’ Intergovernmental Panel on Climate Change. The panel’s mandate will be to gain insights into AI technology and formulate a “State of AI Science” report.
