OpenAI unveils new cybersecurity model

OpenAI has unveiled a specialized AI model, GPT-5.4-Cyber, designed to enhance defensive cybersecurity capabilities for government and intelligence agencies. The company recently briefed U.S. federal officials and the "Five Eyes" intelligence alliance on the tool, which aims to strengthen information security against evolving digital threats. Access to the new system will initially be restricted to a select group of certified specialists and researchers to ensure controlled implementation. This development highlights the intensifying competition in the sector, as rival firm Anthropic is also reportedly advancing its own AI-driven security initiatives.