Bold claim: AI-driven endpoints are rewriting risk, and traditional security isn’t enough anymore. That’s the core issue this piece tackles, and it’s exactly what makes the story of Palo Alto Networks’ move to acquire Koi so consequential. Here’s a fresh, beginner-friendly rewrite that preserves all the key facts while clarifying and expanding where helpful.
Palo Alto Networks has announced its plan to acquire Koi, a pioneer in Agentic Endpoint Security, with the goal of closing a critical security gap as AI tools become more embedded in daily work. The company frames this as the next essential step in reducing enterprise risk: securing the AI-native ecosystem that now governs how work gets done.
Why this matters now
Traditional security models were designed to stop malicious files and obvious threats. But today’s AI-enabled agents and tools can actively read, write, and move data, often with broad access and fewer guardrails. Attackers are increasingly exploiting agent frameworks—bumbing against authentication checks, using API-based remote code execution, or spoofing agent identities to hijack trusted automation. Meanwhile, the surface area isn’t limited to executables; extensions, plugins, packages, scripts, and model artifacts are increasingly shaping endpoint behavior outside centralized oversight. This rapid, agent-driven shift accelerates risk to machine speed, leaving a notable blind spot in conventional security approaches. That is why a new protection category—Agentic Endpoint Security—has become necessary.
What the acquisition aims to deliver
When the deal closes, Koi’s Agentic Endpoint Security will integrate with Palo Alto Networks’ Prisma AI Security (AIRS), expanding protection for AI-led operations. At the same time, the integration is expected to bolster Cortex XDR’s endpoint security by delivering deeper visibility into the AI attack surface. The combined capabilities are intended to give customers clear visibility and practical control over agent-based tools, allowing safer deployment of AI-driven workflows.
Executive perspectives
Lee Klarich, Palo Alto Networks’ Chief Product & Technology Officer, emphasizes that AI agents and tools act as insider threats: they operate with extensive system access but remain largely invisible to traditional security controls. He believes the Koi acquisition will close this visibility gap and set a new standard for endpoint security. The goal is to enable customers to govern every agent, plugin, and script with confidence, ensuring secure usage of AI within protected boundaries.
Amit Assaraf, CEO and Co-Founder of Koi, notes that their mission is to secure the “next frontier” of risk. In an environment dominated by agentic interactions, legacy solutions are often blind. By joining Palo Alto Networks, Koi aims to scale its technology to large enterprises, delivering protection that makes modern AI-native endpoints secure by design.
What this means for customers and the market
The proposed combination envisions a more comprehensive safeguard for AI-powered endpoints, spanning visibility, governance, and security policy enforcement. For organizations, this could translate into safer experimentation with AI tools and faster, more confident adoption of AI-enabled processes without opening new, unmanaged risk pathways.
Controversial angles and questions for the audience
- Is centralizing control over agent-based tools compatible with the speed and autonomy that AI agents promise, or might it slow innovation?
- Does deeper visibility into AI attack surfaces create meaningful benefits for most organizations, or is it primarily valuable to large enterprises with complex ecosystems?
- Could this approach set a new industry standard that others must follow, or will it spark counter-movements toward more decentralized or opt-in agent security models?
If you’re evaluating this for your organization, consider how your current security stack handles agent-based tools, what governance controls you’d need for plugins and scripts, and how you’d measure the effectiveness of Agentic Endpoint Security in reducing real-world risk.
Learn more about Palo Alto Networks’ intent to acquire Koi at their official blog post on securing the agentic endpoint.
Investor-focused details
Palo Alto Networks plans to share additional specifics during its Q2 FY2026 earnings call, scheduled for February 17, 2026, at 1:30 p.m. PT. A live webcast will be available in the Investors section of their site.
About the company
Palo Alto Networks is positioned as a global AI cybersecurity leader, offering a broad portfolio across network, cloud, security operations, AI, and identity. Serving thousands of customers worldwide, the company emphasizes AI-driven platforms designed to simplify security and accelerate safe innovation.
Forward-looking statements and risk note
As with any such announcement, several uncertainties could affect outcomes, including regulatory approvals, successful integration, and the ability to realize projected synergies. The company emphasizes that actual results may differ from projections due to a range of factors, including market dynamics, product development timelines, and customer demand shifts.
Bottom line: a strategic pivot to secure the AI-native endpoint could redefine how organizations protect themselves as work goes increasingly AI-powered. But it also invites debate: will this move meaningfully reduce risk across diverse IT environments, or will it provoke new questions about control, flexibility, and perspective on what “secure” means in an AI-first era?
