LONDON / RankWire.AI / – Britain is ramping up its defenses against the threats posed by advanced AI after discovering more than 1,300 cyber vulnerabilities during recent testing. These security gaps were found across hundreds of government software repositories. The British government explained that its initiative will leverage frontier AI tools to enhance cybersecurity and uncover weaknesses within vital digital infrastructure. Additionally, officials are devising plans to address serious incidents involving autonomous AI systems. This effort is part of a broader strategy to bolster national resilience as increasingly capable AI models are introduced into public use.
In two government-led pilots, a total of 1,352 vulnerabilities across 548 code repositories in 18 departments were identified and remediated. The trials aimed to evaluate how advanced AI could assist in pinpointing software flaws and recognizing potential cyber threats. Authorities now intend to expand the frontier AI hardening initiative throughout government agencies. Particular attention will be given to systems carrying the greatest risk. The program also seeks to develop broader government capabilities in deploying AI for cyber defense, all while ensuring appropriate oversight of the technology.
The Artificial Intelligence Minister Kanishka Narayan stated that Britain’s AI Security Institute currently assesses advanced models and evaluates their safety measures. The institute employs more than 100 technical specialists dedicated to model evaluation and security. Its ongoing work includes testing Gemini 4 Argon and other sophisticated systems. The government emphasizes that independent testing remains essential, as public authorities cannot solely depend on AI developers to assess risks associated with increasingly advanced models.
Expansion of AI incident preparedness
Collaborations between the National Cyber Security Centre and government departments, alongside the AI Security Institute, are underway to establish protocols for major incidents involving agentic AI. These systems have the capacity to perform complex tasks with minimal human oversight. The proposed framework will assist authorities in incident identification, risk mitigation, and supporting those affected. Furthermore, officials are reviewing how current cyber response strategies can be adapted to handle failures involving autonomous AI and advanced technologies within critical service sectors.
Discussions are also ongoing regarding amendments to the Cyber Security and Resilience Bill as it advances through Parliament. The proposed changes include heightened requirements for addressing AI-related cyber risks, serious incidents, near misses, and loss-of-control scenarios. Present proposals mandate that regulated operators and digital service providers report significant cyber incidents promptly, with initial notifications due within 24 hours and comprehensive reports within 72 hours. The security framework encompasses essential sectors such as hospitals, energy grids, and water supply networks.
Enhanced AI safeguards in development
Narayan emphasized that Britain needs to combine robust cyber defenses with independent testing and domestic technical expertise. He also indicated that legislation might be necessary where current safeguards do not adequately mitigate emerging risks. The government has begun creating support mechanisms for AI verification and intends to share further details in the upcoming months. Such verification tools are designed to verify whether advanced systems operate as intended and whether safety controls remain functional under challenging conditions.
These initiatives are part of a broader government effort to examine the implications of artificial intelligence on national security and public services. The Prime Minister’s AI Taskforce is actively reviewing issues related to economic development, public sector performance, workforce impacts, and security concerns. Additionally, Britain has recognized AI as a key policy priority for its 2027 G20 presidency. The latest cybersecurity efforts are primarily focused on testing cutting-edge models, fixing software vulnerabilities, and preparing government systems for potential AI-related crises.