By now, you’ve heard about the latest frontier AI models that are remarkably good at finding vulnerabilities in code and creating potential exploits. So good, in fact, that these models have been significantly limited from general use in an attempt to give defenders time to find and fix vulnerabilities before attackers find and exploit them.

For context, on April 7, 2026, we began testing Anthropic’s Claude Mythos model as a launch partner for Project Glasswing. Our conclusion was clear: The latest models are extraordinarily capable at finding vulnerabilities and changing them into critical exploit paths in near-real-time. In Defender's Guide to the Frontier AI Impact on Cybersecurity, I shared our early findings and recommendations.

Since then, we’ve continued testing the latest frontier AI models, including Anthropic’s Mythos and Claude Opus 4.7 and OpenAI’s GPT-5.5-Cyber as part of the Trusted Access for Cyber program. The big question just a few weeks ago was: “Are we overstating the model capabilities?” With more testing, I can confidently say we weren’t. In fact, these models are likely even better at finding vulnerabilities than we initially realized. Today, we’re providing an update on our ongoing research, our learnings uncovered in the process, and the approach we’re taking to protect our customers.

Find and Fix Before Attackers Find and Exploit

Today, we released our May “Patch Wednesday” security advisories, our monthly cadence of transparent vulnerability disclosure and remediation. This is the first time where the majority of findings were the result of frontier AI models scanning our code.

• These are the results of the full, initial scan of over 130 products across all three platforms.
• As of today, we’ve patched all important vulnerabilities in our SaaS delivered products, and all customer-operated products now have patches available.
• Today’s advisory covers 26 CVEs (representing 75 issues) versus our usual volume (typically less than 5 CVEs in a month); none of which are being exploited in the wild. Note, this excludes CyberArk vulnerabilities, which are disclosed in their normal process.

It's important to understand this isn’t a one-and-done situation. We’re now rescanning, applying all our learnings about how to provide the right context and threat intelligence to the models. We intend to fix every vulnerability we find before advanced AI capabilities become widely available to adversaries.

While incredibly powerful, AI models aren’t simply magic. To achieve high-fidelity results, you need to build AI scanning harnesses, leverage context, guardrails and threat intelligence. We’ve also discovered a variance across models, due to variations in their training. A multimodel approach is required to identify the superset of vulnerabilities. And finally, while the immediate priority is finding and fixing the vulnerabilities that organizations currently have, the longer-term shift is incorporating these models directly into the software development lifecycle. This is the light at the end of the tunnel: A future where software is secure by design.

Four Steps Every Organization Needs to Take Immediately

Regardless of the current restricted access, we believe these capabilities will flow more broadly to other models. We now estimate a narrow three-to-five-month window for organizations to outpace the adversary before AI-driven exploits start to become the new norm. This impending vulnerability deluge demands urgency. Organizations that haven’t put appropriate safeguards in place will face an entirely new class of risk. Here’s what we recommend:

1. Find and Fix Vulnerabilities In Your Applications, Products and Code
   Find and fix before attackers find and exploit.
   • Leverage AI models to identify vulnerabilities across all codebase.
   • Apply the same AI scanning to your open-source supply chain, and remediate or mitigate findings.
   • Run accelerated patching tightly coordinated with product and development teams.
2. Assess, Reduce and Remediate Your Exposure
   Reduce what is reachable by attackers, secure what must be accessible, such as customer-facing applications.
   • Attack surface management products, like Cortex Xpanse^®, have never been more critical for finding and reducing exposure.
   • The latest frontier AI models are very adept (with the right AI scanning harness) at evaluating exposures, understanding security misconfigurations and prioritizing attack-path reachability.
   • Audit your supply chain, including AI infrastructure, runtime environments and model dependencies.
3. Ensure Attack Protections
   Vulnerability exploits are typically just one step of a multi-step attack lifecycle. Ensuring best-in-class protections is now even more important for preventing breaches.
   • Map current sensor coverage to identify critical blind spots in detection, prevention and telemetry.
   • Deploy best-in-class XDR everywhere with an emphasis on real-time ML-based detection and prevention of attacks with all hosts on-premises and cloud included.
   • Deploy Agentic Endpoint Security to secure wide-scale adoption of vibe coding and AI security across the enterprise (e.g. Prisma AIRS^® and our recent acquisition of Koi are now a necessity for securing the agentic endpoint).
   • Secure enterprise browsers with AI-based security are a must have for securing where users now do their work.
   • Zero trust and Identity Security are foundational to securing every user and connection, extending to internal segmentation and outbound application connections.
4. Deploy Real-Time Security Operations
   Autonomous AI-driven attacks will drive attack lifecycles to minutes requiring every SOC to achieve single-digit mean time to detect (MTTD) and mean time to respond (MTTR).
   • Attack detections must be AI/ML-driven to detect even frequently changing and novel attacks at scale.
   • These AI detections must operate against a wide range of first party and third party data sources. A best in class AI SOC must operate on ALL relevant data sources.
   • Automation, both natively integrated and throughout the SOC lifecycle, is necessary to achieve single-digit MTTR. This automation will increasingly be agentic.
   • This must be delivered as a platform to remove seams and gaps created by point solutions.
   • Assess and act as quickly as possible.

Fighting AI with AI — AI Frontier Security Innovations Coming Soon

So far, frontier AI models only find new attacks, not new attack techniques. This means that with the right innovations, we can expand our use of AI to solve the security challenges that organizations are facing, and deliver what our customers need to stay ahead of the ever-evolving threat landscape, including:

• Reimagining virtual patching with proactive, high-fidelity content updates across network, endpoint and cloud security – We expect that across open source and technology suppliers there will be a deluge of patches, and virtual patching will provide a mitigation layer necessary to give your teams time to update. We expect to roll out the first phase of capabilities very soon.
• Enhanced attack preventions, including cyber-LLM trained ML and small language models (SML) and behavior protections – Early testing with Cortex XDR^® and our network security security services, such as WildFire^® malware prevention, indicate high protection coverage from the types of attacks created using these new frontier AI models.
• Using these models to scan our code, applications and even security configurations – Our intention is to productize these capabilities and incorporate them into our platforms.

Unit 42 — We’re Here to Help

We recognize that not everyone has the capacity and/or expertise to action all of the recommendations to effectively counter frontier AI-driven risks in the short timeframe mandated by AI innovation. Our Unit 42 Frontier AI Defense service is designed to discover and remediate your current exposure before attackers do, strengthen controls that reduce exposure and contain impact and modernize security operations so teams can detect and respond at machine speed.

This is a pivotal moment for our industry. While the scale of the challenge presented is real, I’m confident in our ability to solve it. We’re here to help our customers navigate this transition and ensure that as the landscape continues to evolve, the advantage remains with the defender.

Forward-Looking Statements

This blog contains forward-looking statements that involve risks, uncertainties and assumptions, including, without limitation, statements regarding the benefits, impact, or performance or potential benefits, impact or performance of our products and technologies or future products and technologies. These forward-looking statements are not guarantees of future performance, and there are a significant number of factors that could cause actual results to differ materially from statements made in this blog. We identify certain important risks and uncertainties that could affect our results and performance in our most recent Annual Report on Form 10-K, our most recent Quarterly Report on Form 10-Q, and our other filings with the U.S. Securities and Exchange Commission from time-to-time, each of which are available on our website at investors.paloaltonetworks.com and on the SEC's website at www.sec.gov. All forward-looking statements in this blog are based on information available to us as of the date hereof, and we do not assume any obligation to update the forward-looking statements provided to reflect events that occur or circumstances that exist after the date on which they were made.

The post Defender's Guide to the Frontier AI Impact on Cybersecurity: May 2026 Update appeared first on Palo Alto Networks Blog.

The release of the newest frontier AI models marks a turning point for cybersecurity. Palo Alto Networks has conducted early testing of the latest frontier AI models, including Anthropic’s Mythos model as part of Project Glasswing and OpenAI’s latest models as part of Trusted Access for Cyber program. The conclusion is clear: They are extraordinarily capable at finding vulnerabilities and generating corresponding exploits.

This generational improvement in coding ability directly translates to a significant advance in vulnerability discovery and exploit generation. These capabilities, however guardrailed, will not stay contained. Similar advances will appear across other major AI labs, Chinese models, and open source models. Attackers will find the seams in those guardrails. They will use advanced AI to discover zero-day vulnerabilities at scale, generate exploits in near real time, and develop autonomous attack agents unlike anything the industry has faced.

Within six months, advanced AI models with deep cybersecurity capabilities will become commonplace. Organizations that have not put appropriate safeguards in place will face an entirely new class of risk across their enterprise and critical infrastructure.

Frontier AI: A Quantum Leap in Code Fluency

As you have probably already seen, the latest unbounded models like Mythos represent roughly a 50% improvement in coding efficiency over Anthropic’s previous leading model. Palo Alto Networks has had early access to unbounded models and we’ve been able to leverage this vast improvement in coding to a quantum leap in scanning and offensive capability.

Hundreds of our best security engineers have been assessing these capabilities and developing best practices for using it effectively. The results revealed several core truths:

• Vulnerability discovery at scale: Frontier AI is exceptionally effective at identifying vulnerabilities in code. In less than three weeks, it accomplished the equivalent of a full year’s worth of penetration testing effort.
• Attack path determination: Perhaps more impressive than finding individual vulnerabilities, Frontier AI excels at vulnerability chaining, combining multiple lower-severity issues into critical-level exploit paths. For example, linking two medium-severity and one low-severity vulnerability into a single critical exploit.
• Full-stack logic analysis: Frontier AI can analyze the full exposure surface of applications, including SaaS and public-facing platforms, identifying logic-based vulnerabilities that traditional tools miss.

Impacts on the Cyber Landscape

Attackers have been using LLMs for years, but based on our testing of frontier AI models, there are three key areas where they will have a significant impact on the cybersecurity landscape:

1. The Vulnerability Deluge: Frontier AI models will dramatically accelerate the rate at which vulnerabilities are discovered, by defenders and attackers alike. This will be particularly acute in open source and critically, the flood of patches that follows will itself create risk. Every patch that is not applied immediately becomes a known, targetable vulnerability. Organizations will need to accelerate and automate their patching programs, rethink how they prioritize and apply patches, and ensure best-in-class protections are in place to mitigate vulnerability until they can be remediated.
2. Rise of Inside-Out Attacks: Recent supply chain attacks on tools like LiteLLM and Trivy demonstrate a growing pattern where attacks land adversaries inside an organization’s infrastructure, bypassing multiple conventional attack steps and reducing the number of prevention opportunities available to defenders. The rapid deployment of AI infrastructure has made this problem more acute as the AI supply chain, including runtime environments, communication infrastructure, and model dependencies, is often insufficiently protected. While open source usage and patching practices must become significantly more robust, organizations will need structural containment of potential attacks through zero trust, identity modernization, outbound connection restrictions and lateral movement protections.
3. Faster AI-Assisted Attack Cycles: I expect the most consequential shift with frontier AI models is the move from AI-assisted to AI-driven attacks. Attackers will build autonomous attack agents that dramatically compress attack cycle times. What once took days or weeks of skilled manual effort will soon be executed in minutes. This democratization of advanced attack capabilities means that defenders must match that speed with near-real-time detection and response, which is only possible with extensive AI and automation throughout security operations. Organizations whose Mean Time to Detection and Mean Time to Response are not measured in low single-digit minutes will be outpaced.

The Defenders Guide: Assessment, Protection, Platformization

The framework for defending against AI-driven threats is not completely new, but the standard for execution must be absolute. Organizations that are “mostly protected” are effectively unprotected. What follows is a phased approach – assessment, protection and platformization – that organizations should pursue in parallel to close gaps before attackers exploit them.

Assessment: Every organization should use the latest AI models to assess its entire code and application landscape and build a comprehensive asset and exposure inventory.

Key priorities:

• Leverage AI models to identify vulnerabilities across your codebase, applications and infrastructure before attackers do.
• Evaluate exposure with full context, including how vulnerabilities chain together to form critical exploit paths.
• Audit your open source supply chain, including AI infrastructure, runtime environments and model dependencies.
• Map your current sensor coverage. Detection, prevention and telemetry gaps represent critical blind spots.

Protect & Remediation: Remediating and reducing exposure is table-stakes. What in the past may have been difficult due to cross-organizational friction of finding and fixing at pace should now be accelerated with the c-suite attention of these new AI models. But this must go further and extend to comprehensive deployment of best-in-class attack prevention capabilities where the new standard is 100% coverage and optimization.

• XDR everywhere, with emphasis on real-time ML-based detection and prevention of attacks; all hosts on prem and cloud included.
• Agentic endpoint security to secure wide-scale adoption of vibe coding and AI security across the enterprise (e.g. Prisma AIRS and our recent acquisition of Koi is now a necessity for securing the agentic endpoint).
• With an average of 85% of work now happening in the browser, secure enterprise browsers with real-time security become a must-have for attack prevention.
• Zero trust and identity security are foundational to securing every user and every connection.

Real-Time Security Operations: With attack cycle times shrinking rapidly, the legacy approach to security operations simply doesn’t work. Disparate tools analyzing data in silos overlaid with manual processes must be replaced with AI and automation throughout. Cortex XSIAM, our AI-driven SOC platform, is what I consider to be the gold standard for how to take a next-generation approach to deliver MTTD and MTTR in single digit minutes.

• Attack detections must be AI/ML driven to detect even frequently-changing and novel attacks at scale.
• These AI detections must operate against a wide range of 1st party and 3rd party data sources – a best in class AI SOC must operate on ALL relevant data sources.
• Automation both natively integrated and throughout the SOC lifecycle is necessary to achieve single digit MTTR; this automation will increasingly be agentic.
• This must be delivered as a platform to remove the seams and gaps between point solutions.

We’re Here to Help

Achieving this level of resilience requires the right platforms and the right expertise.

To help you navigate this shift, we are introducing Unit 42 Frontier AI Defense. This new offering is designed to discover and remediate your current exposure before attackers do, strengthen controls that reduce exposure and contain impact and modernize operations so teams can detect and respond at machine speed.

This is the moment we’ve been preparing for. The threat has never been more sophisticated, but the path forward has never been clearer, and we’re here to partner with you on what comes next.

The post Defender's Guide to the Frontier AI Impact on Cybersecurity appeared first on Palo Alto Networks Blog.