Back to News
ProductAI Understanding briefing

Palo Alto Networks extends AI firewall to Nvidia’s Open Agent Safety Platform

Palo Alto Networks announced that its Prisma AIR S AI Runtime firewall will integrate Nvidia’s newly launched Open Agent Safety Platform, leveraging the Sentry design and planning a link to Nvidia’s OpenShell runtime.

4 min readRead the linked source
Source-provided image accompanying Palo Alto Networks extends AI firewall to Nvidia’s Open Agent Safety Platform
Source referenceSource recorded
Publisher
tikr.com
Source link
tikr.comhttps://www.tikr.com/blog/palo-alto-networks-stock-rose-nearly-5-as-nvidia-launched-an-ai-agent-safety-platform-heres-what-palo-alto-is-building-on-top-of-it?
Source type
Linked source — primary-source status has not been established.
ContextUnderstand this in 60 seconds

Start here

Key terms

Embedding
A numeric vector representation that captures semantic meaning of text, images, or other data.
Latency
The time between sending a request and receiving the model's output.
Test yourselfAI Agents Quiz

What happened

Palo Alto Networks said its AI firewall, Prisma AIR S, now runs natively on Nvidia’s BlueField hardware and incorporates the Sentry component of Nvidia’s Open Agent Safety Platform. The company also outlined a future integration of its Idira identity suite with Nvidia’s OpenShell runtime.

In a blog post published on September 28, Palo Alto Networks detailed how its Prisma AIR S AI Runtime firewall, which already runs on Nvidia’s BlueField‑4 SmartNIC, will now draw on the Sentry component of Nvidia’s Open Agent Safety Platform. The company described the integration as an "AI gateway" built on Nvidia’s Vera CPU architecture.

Palo Alto also mentioned a planned connection between its Idira identity business—formerly CyberArk—and Nvidia’s OpenShell runtime, though it did not disclose any revenue expectations for these components.

The announcement coincided with Nvidia’s launch of the Open Agent Safety Platform, a suite of tools designed to curb rogue AI‑agent behavior. Palo Alto’s move positions it among the first security vendors to embed its firewall directly into the AI‑agent safety stack.

Source details: tikr.com ↗

Why it matters

The integration shows a growing convergence between network security vendors and AI‑agent safety tooling, indicating that AI‑driven workloads will increasingly rely on specialized security layers built directly into accelerator hardware. By its firewall in Nvidia’s stack, Palo Alto aims to protect the “front‑end network traffic of the AI factory,” a segment that traditional firewalls have struggled to secure. This move could set a precedent for other security firms to offer hardware‑native AI‑agent protections, potentially reshaping how data‑center operators secure multi‑tenant AI workloads.

security functions into the same silicon that runs AI agents reduces and attack surface compared with traditional, separate firewall appliances. This could become a critical differentiator as AI workloads scale and as incidents like the September 20 OpenAI sandbox escape raise concerns about agent misbehavior.

Palo Alto’s strategy reflects CEO Nikesh Arora’s view that most data‑center capacity is not multi‑tenant, limiting opportunities for conventional firewalls. By targeting the AI‑factory layer, the company hopes to capture a niche where hyperscalers and AI developers need built‑in safeguards.

If successful, this model may encourage broader adoption of hardware‑native security solutions, prompting cloud providers and AI platform vendors to prioritize integrated safety features in their product roadmaps.

Interactive Mechanism

Interactive Mechanism: How It Actually Works

Explore the underlying technology behind this development interactively.

Agent Lifecycle Stage:
1
User Intent & Planning: "Audit customer refund request #4092 and settle payment."
2
Tool Calling: Emits structured JSON call crm_get_transaction(id='4092').
3
Guardrail & Verification:🛡️ Paused: High-value action requires human operator sign-off.
4
Final Settlement: Refund recorded, email receipt dispatched, and audit log stored.
Core takeaway: An AI agent is not just a language model—it is a closed loop of planning, tool invocation, and environment feedback. Production systems require self-healing retries and strict human approval guardrails.
Interactive Concept Check+10 Points
AI Agents Quiz

What most distinguishes an AI agent from a basic chatbot?

What to watch next

Investors and analysts will watch Palo Alto’s upcoming fiscal Q1 earnings for any disclosed revenue tied to the Nvidia‑based offering, and monitor whether other security vendors announce similar hardware‑native AI‑agent solutions.

Palo Alto’s fiscal first‑quarter earnings, due later this year, will reveal whether the Nvidia‑based products generate measurable revenue and how the market responds to the integration.

Analysts will track whether competitors such as CrowdStrike or Zscaler announce comparable hardware‑native AI‑agent security offerings.

Further developments from Nvidia, such as expanded OpenShell capabilities or additional watchdog features, could enhance the value proposition of Palo Alto’s integrated firewall.

Related guides & quizzes

AI AgentsAI Models ExplainedFuture of AIAI EthicsTest what you know — try a free AI quizLook up an AI term in our glossaryFollow the AI model release tracker
Found this useful?