Technology companies are increasingly turning to AI systems to defend against AI-powered attacks.
IBM said it has begun integrating AI systems into vulnerability analysis, remediation prioritization, software testing and response coordination. The company also said it is using AI code assistant to automate some of its testing and patch management workflows.
“IBM takes a multi-model approach to security,” Thomas wrote in a blog post. “We applied AI models, including frontier models such as Claude’s Mythos Preview, to these functions across the defense.”
The Myth of Claude belongs to a new category of AI systems designed for cybersecurity analysis. Security-focused models help researchers identify vulnerabilities, analyze malware, and simulate attacks on corporate environments.
Hughes said AI has dramatically shortened the time between vulnerability disclosure and exploitation.
“The time is [it took] A year ago, it took an average of 23 days to exploit a published vulnerability; today, using some of these frontier models, we’re down to 9 hours. So this is real. ”
IBM said that as attacks become more sophisticated, companies need to further automate their vulnerability response processes. The company also pointed to its Zero Trust security system, which continuously verifies users and devices, rather than assuming that everything in the network is automatically secure.
Another part of the strategy includes IBM Autonomous Security, a platform that uses AI agents to automate some of the threat detection and response.
“It’s not just a matter of waiting for a patch to come out,” Hughes said. “A lot of the response here is about making existing security controls work much more effectively and much faster than they did before.”
