JumpCloud

AI is touching more of your environment than ever - ITSM, CRM, financial systems, IAM. Organizations report AI has access to an average of 3 types of company systems or data. But 72% see a moderate-to-large gap between what AI is permitted to access and what they can actually prove it accessed. Access without proof isn't control. It's exposure. See how identity, permissions, and evidence need to work together to keep pace with where AI is already operating.

Your AI agents can change systems, adjust permissions, and modify records. But 59% of organizations using agents haven't fully integrated them into the IAM policies they already trust for human users. When an agent acts, you need to know which identity was behind it, what it was authorized to do, and who owns that accountability. Right now, most organizations can't connect all of those pieces. See how 250 IT decision-makers are approaching agent identity governance, and where the gaps are most consequential.

Most AI research out there right now will tell you AI adoption is up. That's true, but it's only half the story. JumpCloud's new research, out now, reveals that confidence in AI maturity is falling even as deployment grows. Six months ago, 40% of IT leaders described their organizations as mature in AI deployment. Today that number is 23%. Not because organizations are pulling back from AI, but because the ones actually running agents in production are getting honest about what governing them requires. That tension is what we built this report around. Not how many organizations are deploying AI, but whether they have the foundations to do it safely and at scale. A few things that stood out to me personally: organizations in fully unified IT environments are five times more likely to deploy agents in business-critical workflows than those running fragmented stacks. Non-human identities now outnumber humans in 83% of organizations, yet only 21% have governance controls in place for them. And the organizations furthest along in AI execution are five times more likely to report no barriers to expanding AI at all. The gap between AI adoption and AI readiness is the defining business challenge of the next few years. The organizations that close it first won't just run AI more securely. They'll move faster, spend smarter, and say yes to more. You can download the full report here: https://lnkd.in/ghkzdaQr Would love to hear what resonates with your teams.

10 years. 52 million monthly users. One identity platform. Proud to share how Grab, the superapp powering everyday life across Southeast Asia, has scaled its security with JumpCloud since 2016. What started as a move away from on-prem Active Directory has grown into a fully programmable identity platform: dynamic groups for zero-touch onboarding, passwordless access with JumpCloud Go™, and API-driven automation for MFA and credential rotation. As Ken Wong, Grab's Head of Technology Solutions, put it: JumpCloud has been their "quiet engine for ten years" - from solving admin chaos in 2016 to enabling a passwordless future today. This is what it looks like when identity becomes an engineering discipline, not just an IT function. Huge congratulations to the Grab team on a decade of secure, borderless growth. Read the full case study: https://lnkd.in/ej8KfeN3

Companies should be able to move fast with AI without losing visibility into what’s happening across the business. That gets harder as adoption spreads. Developers are connecting MCP servers to tools like Cursor and Claude Code, teams are experimenting with new AI services, and usage is growing across employee devices. Most of that is happening for a good reason: people have work to get done, and these tools help them do it faster. But it also means IT can quickly find itself trying to govern an AI environment it can’t fully see. We’ve been thinking a lot about how to close that gap without putting more friction in the way of employees. JumpCloud can now discover public MCP connections and track AI token consumption directly from managed devices, giving IT a clearer view of how AI is being used, what it’s connecting to, and what that usage is costing (without having to build separate API integrations for every tool). To me, that’s what Intelligent, secure IT should enable: giving people the freedom to use great technology while giving IT the visibility and control to manage it responsibly. More from the team here: https://lnkd.in/gQgG4Hpg

If we’re going to ask customers to trust JumpCloud to govern their AI agents, we should be willing to trust it with ours. That was a big part of the thinking behind becoming Customer Zero for Agentic IAM. Like a lot of companies, we already have AI agents being built and used across JumpCloud for a growing range of use cases. We’re bringing those agents through our own Agentic IAM platform so they have an identity, a named human owner, deliberately scoped access, visibility into their activity, and a lifecycle IT and Security can manage. We can have a point of view on how companies should govern agents. But there’s a big difference between having a point of view and actually doing it across your own organization as adoption grows. And doing it across many different use cases is where things get interesting. Agents will have different owners, interact with different systems, require different levels of access, and become more capable over time. Managing that in our own environment gives us firsthand experience with the same challenges our customers are beginning to navigate. We’ll find friction. We’ll find gaps. Some of what we learn will influence what we build. It should. You earn trust by being willing to use what you build in the same real-world conditions as the people you’re building it for. For us, being Customer Zero means putting Agentic IAM to work across our own company, learning from it, and building a better product as a result. Here’s a look at what we’re learning: https://lnkd.in/duXqWuyi