UK Financial Watchdogs Warn of Bank Security Vulnerabilities in AI Era

UK financial authorities have warned that the advancement of artificial intelligence (AI) technology is leading to an increase in security vulnerabilities within banking systems and a greater potential for criminal exploitation. The head of the UK's Prudential Regulation Authority (PRA) identified AI-related security issues as the most serious risk factor that could shake the stability of the entire financial sector in the future. Chief Executive Officer Sam Woods cited Anthropic's AI model 'Claude-3' as an example, explaining that the emergence of technology capable of finding cybersecurity vulnerabilities at an expert level is exacerbating the risks. As Anthropic expands access to its AI models, Woods indicated that some UK banks could be included. He diagnosed that while the financial sector can leverage AI to enhance its security capabilities, it simultaneously faces new threats.

CEO Sam Woods proposed concrete countermeasures, including expanding the speed and scope of software patches to fix security vulnerabilities, identifying the use of high-risk open-source code, and prioritizing security reinforcement projects over new service development, suggesting a reordering of IT investment priorities.

Woods stated that the pace of AI technology development is so rapid that it is unrealistic for supervisory authorities to individually regulate banks' use of AI. He clarified that the role of financial authorities should not be to block the use of technology, but rather to continuously monitor its impact on financial stability and security, and to focus on controlling key risks.