Question to the Cabinet Office:
To ask His Majesty's Government what assessment they have made of work undertaken by AI companies in the United Kingdom on recursive self-improvement; and whether such work could lead to the development of superintelligent AI in the United Kingdom.
The AI Security Institute (AISI) was established to build a rigorous, scientific understanding of the capabilities of the most advanced AI systems and the risks they pose. It generates evidence about the capabilities of models at the frontier, and how fast they are changing, to inform national security and public safety. AISI is working at pace with industry and international parters to build our understanding of recursive self-improvement and its potential implications for AI security.