Palo Alto AI Cybersecurity: Rival Models Join the Defense

Palo Alto Networks said Tuesday it will launch a new cybersecurity service for businesses built on AI models from Anthropic and OpenAI, putting two rival AI labs' technology to work in the same defensive product. The Palo Alto AI cybersecurity offering will hunt for vulnerabilities across corporate systems, with models doing the probing and analysis that used to take human security teams weeks.

The announcement, reported by Reuters, puts Palo Alto Networks in a growing crowd of security firms betting that AI-on-AI is the only way to keep up. Attackers increasingly use AI to find and exploit weaknesses in corporate networks, so the company wants defenses that can detect and respond to threats at machine speed.

What the service will do

The service is called Unit 42 Continuous Frontier AI Defense. It will use cyber-focused AI models including Anthropic's Claude Mythos 5 and OpenAI's GPT-5.6-Cyber, alongside open-weight models, according to Reuters. The use of both companies' flagship models in one product is a notable choice, given that Anthropic and OpenAI compete directly for the same enterprise customers.

Rather than running a single audit, the service is built to test continuously. It will check web applications, application programming interfaces, and cloud infrastructure on an ongoing basis, flagging vulnerabilities and mapping possible attack paths as a company's digital environment changes. The idea is that a network never sits still: software gets updated, new services get deployed, and configurations drift, so a point-in-time security review is already out of date by the time it is delivered.

When the system finds a problem, it will not just report it. Palo Alto said the service will also provide guidance on fixing security gaps, including code-level fixes and virtual patching options. Virtual patching is a stopgap that blocks an attack without taking a system offline, which matters for businesses that cannot afford downtime while a full software fix is developed and tested.

Why the timing makes sense

The shift reflects how both sides of the security fight now operate. Automated tools can probe thousands of systems for a single known flaw in the time a human analyst would need to check a handful. For the companies defending those systems, the bottleneck has become human review: there are far more alerts and vulnerabilities than there are people to work through them.

Palo Alto Networks is betting that models built for security work will handle the routine parts of defense better than general-purpose AI. The tasks are concrete rather than creative: reading logs, tracing how an attacker could move through a network, and checking whether a patch closed the hole it was meant to close. Getting those right frees human analysts to focus on the threats that require judgment.

The approach also reflects a practical reality for large companies. Most run a mix of web applications, APIs, and cloud services from different vendors, and each addition widens the surface attackers can probe. A testing service that keeps pace with those changes addresses the gap between the speed of deployment and the speed of manual security review.

How it will be sold

The service will be available globally through annual subscriptions, the company said. Pricing will depend on the mix of OpenAI, Anthropic, and open-source models each customer selects, which gives businesses a way to balance cost and capability. A customer could lean on open-weight models for routine scanning and reserve commercial models for the trickiest investigations, or go the other way if it prefers.

Palo Alto Networks did not say when the service will reach customers or how many businesses have signed up. The company, which trades on the Nasdaq under the ticker PANW, has been expanding its portfolio of AI-assisted security products as the market for automated defense grows.

Whether combining two competing AI labs in one service becomes a pattern will depend on how well the models actually perform at security work. For now, Palo Alto Networks is making a straightforward argument: the attackers have automation, so the defenders need it too.