The U.K. government has taken a significant step towards global AI security standards by releasing voluntary guidance for developers and vendors. The guidance includes recommendations such as monitoring AI system behavior and conducting model testing to protect against hacking and sabotage.
Minister for AI and Intellectual Property Jonathan Camrose emphasized the importance of ensuring confidence in adopting AI technology within organizations in the U.K., especially in the face of a complex cybersecurity landscape. The guidance also calls for strengthening AI supply chain security and minimizing risks from vulnerable AI systems, such as data loss.
Some of the key measures outlined in the guidance include procuring secure software components only from verified third-party developers, training AI developers on secure coding practices, and implementing security guardrails for different AI models. The aim is to promote security by design in AI systems and transform the guidance into a global standard.
To gather feedback and input from stakeholders, the U.K. government has opened a consultation on the guidance document until July 10. This initiative aligns with the government’s commitment to a shared global approach to AI safety, as discussed during a summit held in November.
In addition to the release of the guidance, the U.K. AI Safety Institute recently unveiled an AI model evaluation platform called Inspect, which allows startups, academia, and AI developers to assess the capabilities of individual models and generate a score based on their performance.
Furthermore, collaboration between the U.S. and the U.K. AI Safety Institutes aims to develop safety evaluation mechanisms and guidance for emerging risks in the AI sector. This partnership reflects a broader effort to align on AI safety practices and share resources between the two countries.
