Safety of AI
I agree with you that advanced artificial intelligence systems pose distinct risks. This technology requires trust, which should be a prerequisite for its adoption. To seize AI’s full potential, people have to feel safe and protected for them to confidently explore it.
I believe the majority of AI should be regulated at the point of use by expert regulators. The Government is committed to introducing binding requirements on companies developing the most advanced AI models, to ensure the safe development and use of AI.
Existing regulators are ill-equipped to deal with the dramatic development of new technologies. As such, I am glad to see the establishment of the Regulatory Innovation Office (RIO) and the relaunch of the AI Security Institute. The RIO will help regulators update regulations and to ensure they have the expertise and resources to effectively regulate AI in their respective domains. The AI Security Institute will be taking a leading role in testing AI models regardless of its origin. I also welcome the Government’s commitment to conduct scientifically informed tests to understand new AI capabilities and the risk they pose.
I am pleased that the Government is developing a comprehensive AI legislation which would allow us to safely realise the enormous benefits of these systems. I look forward to the Government’s further engagement with the full range of stakeholders on legislative proposals to establish appropriate oversight.