trending
OpenAI has widened access to its latest cybersecurity-focused artificial intelligence model, GPT-5.5 Cyber, granting several of the UK’s largest banks the ability to deploy the technology as part of their cyber defence operations.
OpenAI expands Cybersecurity AI Access
The move will see institutions including Lloyds Banking Group, HSBC and Nationwide gain access to the model, while NatWest and Santander are already using the technology through existing partnerships. OpenAI is also extending availability to selected financial institutions across Europe, Canada and Japan.
AI Security Race Intensifies
The decision comes amid growing competition between leading AI developers over how advanced cybersecurity tools should be distributed. Rival firm Anthropic has maintained a far more restrictive approach to its Claude Mythos model, limiting access primarily to a small number of US organisations.
The debate has intensified as banks and regulators increasingly explore how artificial intelligence can strengthen resilience against cyber threats. Financial institutions continue to operate highly complex technology estates, often containing decades-old legacy systems that can be difficult to monitor and secure effectively.
Supporting Critical Financial Infrastructure
George Osborne, former UK Chancellor and now a senior executive at OpenAI, said the company believes advanced cybersecurity capabilities should be made available to organisations responsible for protecting critical infrastructure.
OpenAI argues that controlled deployment of AI-powered security tools can help financial institutions identify software vulnerabilities, improve threat detection and accelerate remediation efforts before weaknesses can be exploited by criminals.
Recent testing by the UK’s AI Security Institute found that GPT-5.5 Cyber and Anthropic’s Claude Mythos demonstrated broadly comparable performance across a range of cybersecurity assessment tasks.
Legacy Systems Present Major Opportunity
Cybersecurity specialists believe AI models could prove particularly valuable in analysing older banking systems that contain millions of lines of code. While human expertise remains essential to validate findings and eliminate false positives, AI-driven tools have the potential to dramatically reduce the time required to uncover hidden vulnerabilities.
As cyber threats continue to evolve, banks are increasingly viewing advanced AI as a critical component of future security strategies.
Comments