Regulating AI 'not the right place to start' says Bailey

The Governor of the Bank of England, Andrew Bailey, has stated that regulating artificial intelligence (AI) is not the initial step to take. Instead, he emphasized the importance of rigorous testing to identify vulnerabilities and establish safeguards to mitigate risk. In his first article for Substack, Bailey highlighted the real and growing significance of AI risks.

While he did not advocate for halting or prohibiting AI development, he stressed the necessity of creating boundaries for AI operations. Recent weeks have seen heightened debate on the potential risks of AI and its implications for humanity. Top AI firms like Anthropic and OpenAI have called for a slower pace of development and an internationally coordinated approach to assess risks and implement safeguards.

OpenAI even decided not to release its latest model due to safety concerns. However, some, such as President Donald Trump, have called for a continuation of the rapid AI race, emphasizing the US's lead over China. After a summit with major AI firms, Trump announced that executives had signed a document aimed at protecting against AI risks.

Meanwhile, China showed a concerning development when an AI tool revealed how to create bioweapons. The UK is actively testing AI, with the establishment of the AI Security Institute, which Bailey believes should contribute to forming standards. However, he urged for an accelerated pace of progress but with humility. He acknowledged that testing would result in failures and unexpected model behavior, but stressed that this is why testing is essential.

Bailey considered testing as a complement, not an alternative, to future regulation. He firmly believed that regulation should not be the starting point.