Mimic Virtual Patching Enables Fast, Safe Mitigation of AI-Discovered Vulnerabilities
Powered by Mimic's patented kernel technology, Virtual Patching is the only safe way to quickly and automatically protect against the exploitation of vulnerabilities that frontier models discover, giving enterprises time to fully test a long-term patch.
PALO ALTO, Calif., Nov. 19, 2026 /PRNewswire/ -- In just the past week, personal and corporate email, banking, air travel, and government networks have all been severely disrupted; by attempts to patch serious new vulnerabilities discovered using frontier AI models. Palo Alto Networks (PANW) CEO Nikesh Arora, speaking on the company's earnings call on 18 November 2026, attributed these disruptions to failures of patching processes in the face of a mounting wave of AI-discovered vulnerabilities: "We're seeing critical and severe vulnerabilities discovered nearly seven times the monthly average of last year. These discoveries are happening at a speed and intensity which is hard for IT teams to cope with… We believe many of the major disruptions of the last few weeks, including outages across major enterprises and governments, were partially due to unintended consequences of patching and configuration changes, as IT teams simply cannot patch these vulnerabilities fast enough without compromising their production environments."
In response to this new class of business risks, Mimic has introduced its Virtual Patching offering - a compensating control which rapidly and safely eliminates the risk associated with exploitation of unpatched, newly-discovered vulnerabilities until the full lifecycle of vendor binary patch development, testing, and deployment can be executed.
AI-discovered vulnerabilities create severe business risk
The patch attempts that led to the recent string of outages created severe disruptions in important business services, and large impacts on the providers of those services. The 13 November 2026 nationwide FAA ground stop led to cancellation of more than 20,000 flights in the United States. An online banking system outage at JP Morgan Chase on 16 November 2026 disrupted banking services across a wide range of other banks and financial service providers. A Microsoft 365 outage on 12 November 2026 interrupted email service to millions of Outlook email customers for 24 hours.
Legacy vulnerability- and patch-management tools and processes cannot operate at the scale and pace of AI-powered vulnerability discovery
Speaking with Reuters, Anne Neuberger, who formerly served as deputy National Security Adviser, said "AI is enabling enterprises to find 10 times more vulnerabilities, and when patching at scale, some can cause problems." Yuval Itzchakov, CTO and co-founder of A Security, highlighted the impact of the problem, telling Reuters that "emergency patches go into core infrastructure within hours, often without the full testing cycle, and they are intrusive by design: they change runtimes, kernels and network components that everything else depends on. That speed and scale of change is itself a source of failure".
Binary patching can be fast or safe, but not both
A vendor patch is essentially a new version of the vulnerable application that's being patched. The patch is a new set of binary executable files, with new logic, new file hashes, and a new code signature. Because a patch is a new version of an application, it needs to go through a full version-upgrade lifecycle in order to be safely deployable in a production environment. Every step of this lifecycle needs to be done with the diligence required for a production deployment of a version of the application. There's a limit to how fast this can all be done without cutting corners, and if corners are cut, risks of failure are introduced into the process, and the kinds of service outages described above are the result.
While there's a limit to how fast the patching process can be safely executed, there's no corresponding limit to how fast an unpatched vulnerability can be exploited. That speed disparity creates the need for a new kind of control that allows vulnerable applications to continue running without the risk of exploitation during the window between vulnerability disclosure and the end of the patching cycle. That's what Mimic has built.
Mimic Virtual Patching is the only control that can respond fast and safely to newly-discovered vulnerabilities
Unlike vendor binary patching, Mimic Virtual patching requires no changes to the vulnerable application; the application binary file, its file hash, and its code signature remain unchanged, and no regression or compatibility testing is required.
Mimic Virtual Patches are runtime policy controls which block the execution of exploits while allowing the unchanged application to continue running and performing its business functions normally. A single Mimic Virtual Patch typically mitigates an entire class of vulnerabilities, including vulnerabilities not yet discovered or disclosed. Mimic Virtual Patches can be applied in minutes, without requiring a system or application restart, and can be withdrawn at any time in case application of the Virtual Patch interferes with any normal application or system function.
The Mimic Virtual Patch application workflow can be fully automated, or it can be integrated into an enterprise review and approval process to meet compliance requirements.
About Mimic
Mimic has developed a new class of AI-powered security controls which utilize patented kernel technology to safely and quickly prevent damage from ransomware, from failed change control, from AI Agents gone rogue, and, most importantly, from exploitation of unpatched or previously unknown vulnerabilities discovered by frontier models in minutes or seconds, before they can be exploited.
Learn more: www.mimic.com
Start a free POC: [email protected]
SOURCE Mimic

Share this article