Providing secure access for your human staff is no simple task. For one, traditional VPNs often grant unnecessarily broad access, so you may have replaced that aging technology with a newer zero-trust network access (ZTNA) solution. And you may have hammered your human-centered privileged access management (PAM) tools into working order, so your engineers are just as happy as your security auditors.
But what happens when you add dozens or hundreds of AI agents to the equation? We know that today’s enterprise is racing to become agentic.
Or what if you want to run continuous integration (CI) and continuous deployment (CD) jobs? After all, you want a modern DevOps stack! You’ll learn quickly that the tools you built and tuned for human access come up short.
Checking identity upfront isn’t enough to securely control access. You need to be able to tailor access based on need, not merely identity. And you’ll need to be able to revoke access when work is completed and provide an audit trail to boot. Even more, you don’t want to run separate access architecture for humans and agents.
To support both human and agentic network access, companies need a unified architecture that can handle all access requests without compromise or delay. That means one consistent policy regardless of who requests access or what they need.
Join us on July 22
Today, there are frameworks and tools you can use to harmonize network access control, intelligent PAM, and agentic control into a single access layer that can handle any request from any source at any time.
On July 28, 2026, Tailscale Solutions Engineer Kartik Bharath and Tailscale Product Manager XingLu Wang will join The New Stack to dig into the difficulties that human-first access policies face in the agentic era and how to address them.
We’ll walk you through what a unified access architecture actually covers today: human developers, contractors, pipelines, and AI agents, under one consistent policy model, without adding IT overhead or slowing down engineering teams.
How to join this free, live event
You have successfully registered for the webinar.
What you’ll learn:
- A clear model for extending your existing access controls to non-human identities without needing a separate tool for each identity type
- Architectural guidance on where identity-based networking, PAM, and AI agent policy fit together
- A practical starting point for teams that need to move fast without creating new security debt
The post What happens when your VPN meets 200 AI agents appeared first on The New Stack.
