Bailey Pushes for Rigorous AI Testing Before Regulation
Bank of England Governor Andrew Bailey has emphasized the importance of rigorous AI testing before formal regulation. He believes that policymakers should identify vulnerabilities in AI systems before deciding how to establish a regulatory framework. According to Bailey, 'AI risks are real and increasingly significant,' but regulation is not the right place to start.
Bailey wants to see clear safeguards and credible points for human intervention in AI development. He notes that society should preserve the ability to set boundaries around advanced AI and that testing before and after deployment can expose vulnerabilities and check whether safeguards can contain failures.
The UK has already established the AI Security Institute, which Bailey says is helping build the scientific and technical foundations for AI assurance. However, he adds that 'the pace of progress must accelerate.' Bailey also addresses unexpected model behavior, stating that it does not show testing has failed but instead demonstrates why testing is necessary.