Cisco Systems has unveiled Live Protect, a new platform meant to help customers and businesses defend against AI-enhanced cybersecurity threats without shutting down the systems they rely on every day. The pitch is simple: let security teams move faster than exploits discovered by increasingly powerful AI models, while keeping networks and hardware running.
Dan Howley, speaking during the discussion on Market Domination, said the platform is aimed at an era in which software flaws can be found and used almost automatically. He described Cisco’s approach as a way to click a toggle and put up a shield against a particular exploit, instead of powering down hardware or taking a network offline to install a fix. That matters because some systems only get updated a few times a year, leaving long stretches when they are exposed.
Howley said AI has turned cybersecurity into an arms race. He pointed to Anthropic’s Claude Mythos preview as able to find thousands of vulnerabilities in different pieces of software, a sign of how quickly attackers can scale their reach when machines help do the scanning. Cisco is trying to give defenders a different kind of speed, one that lets them respond without waiting for a full maintenance window.
The problem is that the speed gap may not close so neatly. Howley said the bad guys are always ahead of the good guys, and he also noted that AI coding is pushing teams to publish and push code faster and faster. That combination leaves security teams facing more software, more flaws and less time to respond, even as vendors promise smarter protection.
Cisco’s launch fits into a wider shift in which AI is not only helping companies build software and cloud tools faster, but also giving attackers new ways to probe them. The company is betting that Live Protect can buy time in places where patch cycles are slow and shutdowns are expensive. What it has not yet shown is how well the system will hold up when those attacks move from a demo to a live network.
