LEADING CYBER MODELS
Increased efficacy using gated capability models, including Mythos & GPT-Cyber.
CONTINUOUS DISCOVERY
Frontier AI turns blind spots into attack paths you may see too late.
Search
PREPARE FOR EMERGING RISKS
Continuous Frontier AI Defense
Expert-led, agentic offensive security service that uses exclusive access to gated capability models to automatically discover, validate and remediate exposures before they can be exploited.
MULTI-MODEL HARNESS
Multi-model architecture to align models to their strengths and cover individual model blind spots.
CONTINUOUS ENGINE
Ongoing exposure discovery, testing, and validation to offer near real-time security.
ADVANCED ADVERSARY SIMULATION
Validated end-to-end attack paths to prove real-world exploitability.
ACCELERATED REMEDIATION
Accelerated exposure mitigation and remediation to reduce the risk of breach.
Our intelligence-driven approach
Proprietary technology and ongoing expert guidance shorten the path from risk to protection.
01
SCOPE
We agree on testing scope across your estate.
02
DISCOVER
We execute attacks using gated capability models to map exposures and identify potential weaknesses.
03
VALIDATE
Our offensive security experts verify frontier AI findings and chain end-to-end attack paths.
04
REMEDIATE
Our tailored remediations integrate with your ticketing systems.
05
IMPROVE
We trigger ongoing re-testing automatically as environments or applications change.
Turning Threat Intelligence into Cyber Resilience
Attackers move fast and adapt constantly. Unit 42 brings the experience, technology and operational depth to match that speed. From readiness through response to long-term improvement, we help you act with precision and continuously strengthen your security program.
Fracturing Software Security with Frontier AI Models
Unit 42 recently got hands on with frontier AI models, and our initial findings indicate a major shift in the speed, scale and capability of AI models to identify software vulnerabilities.
Defend Against Frontier AI Risks with Anthropic’s Mythos 5
Unit 42 is expanding its frontier AI services with Anthropic’s Claude Mythos 5, giving organizations access to its advanced cyber capabilities.
Putting OpenAI Cyber Models to Work for Defenders
Through our partnership with OpenAI, we are expanding Unit 42 frontier AI services to put advanced frontier cyber models directly to work in customer environments.
FAQ
Frequently Asked Questions About Frontier AI Defense
Unit 42 is the threat intelligence, incident response, and security consulting arm of Palo Alto Networks helping organizations prepare for, respond to and recover from cyber incidents.
Frontier AI refers to the most advanced AI models, including Anthropic Mythos and OpenAI GPT-Cyber, capable of reasoning across complex tasks, generating code and autonomously executing multistep workflows.
Frontier AI can help attackers identify unknown vulnerabilities, develop exploit code, chain together attack paths and operate at greater speed and scale. Security teams need to account for both AI-enabled attacks against their environment and risks created by their own AI systems, agents and software supply chain.
Unit 42 Frontier AI Defense helps organizations identify and validate exposures before attackers weaponize them. Using frontier AI models and Unit 42 adversary expertise, we simulate advanced attacks, test exploitability and provide evidence-based recommendations on where to focus.
Unit 42 uses a multi-model approach alongside its offensive security experts and real-world threat intelligence. This helps identify more potential exposures, validate which attack paths are truly exploitable and focus recommendations on the risks that matter most.