UK Regulator Says Testing Before Regulation Key to Containing AI Risks

The Governor of the Bank of England has emphasized the need for rigorous testing to identify vulnerabilities and safeguards in artificial intelligence (AI). In an article for Substack, Andrew Bailey stated that regulating AI is not a priority at this stage. Instead, he advocated for a system to intervene and establish boundaries within which AI operates.

Bailey acknowledged that the risks associated with AI are real and significant, but emphasized that halting or prohibiting its development would be counterproductive. He highlighted the immense benefits of AI and stated that there must be a framework in place to assess risks and contain them.

The UK has established the AI Security Institute for testing AI systems, which Bailey believes will contribute to setting standards. However, he cautioned that with accelerated progress comes the risk of failures and unexpected behavior from models.