This is a flawed statement that couldn’t be further from the truth.
Active Directory has survived basically every technology trend that was supposed to make it irrelevant.
The directory wars. Novell NDS/eDirectory was the serious competitor when AD launched. AD won that one outright. OpenLDAP and later FreeIPA found their niches, but neither became the enterprise default for Windows environments.
“The PC is dead.” First thin clients and Citrix, then mobile and BYOD. Phones don’t domain-join. Most Macs don’t either. But AD often still owns the Windows estate those users sit in front of, plus much of the infrastructure behind the VDI session.
Federation was going to make the directory less important. Instead, AD FS became another layer attached to Active Directory. Then cloud identity arrived, and Okta, Ping, and Azure AD/Entra often federated with AD or synchronized identities from it. The authentication experience changed. The directory underneath it often didn’t.
Cloud was going to empty the traditional enterprise datacenter. In 2018, Gartner’s Dave Cappuccio predicted that 80% of enterprises would shut down their traditional datacenters by 2025. Yet on-prem infrastructure is still everywhere. A 2025 poll of 246 IT pros conducted by a principal engineer on Microsoft’s AD team, and later reported by IS Decisions, found 93% describing their environments as hybrid, 54% still running on-prem workloads on AD, and 36% saying they would never be cloud-only.
Azure AD / Entra was going to replace AD. Instead, Entra became Microsoft’s modern cloud identity platform. It did not simply move Active Directory to the cloud. It doesn’t replace Kerberos, NTLM, LDAP, Group Policy, domain join, or machine accounts. Microsoft literally has a separate product—Entra Domain Services—to provide those traditional capabilities for workloads that still need them.
Zero Trust and “identity is the new perimeter.” That didn’t kill AD either. It layered modern controls like MFA, Conditional Access, device trust, and identity protection around increasingly hybrid identity environments. Meanwhile, AD often remained underneath, handling Windows authentication and authorization.
Security failures were supposed to force a rip-and-replace. Mimikatz. Golden Tickets. NotPetya. Years of ransomware crews abusing Active Directory.
And yet Active Directory still hasn’t been retired.
Instead we moved to implement: tiering, PAWs, LAPS, protected users group, fine-grained password policies, logon restrictions, restricting/removing legacy protocols like RC4 and LLMNR and NTLMv1, adding better/additional monitoring, and much more.
As recently as December 2025, Microsoft was still describing AD DS as “central to enterprise identity.”
Containers and “servers don’t need a domain anymore.” And that’s true for a lot of new workloads. But plenty of ERP systems, file servers, SQL deployments using Windows authentication, print infrastructure, NPS, AD CS, and legacy applications still depend on Active Directory.
Those dependencies are what many organizations continue to rely on.
Now we’re entering the AI era.
This one will be incredibly difficult for Active Directory to contend with too. AI agents, non-human identities, autonomous workflows, SaaS everywhere, and entirely new ways of thinking about authentication and authorization will put enormous pressure on a directory designed in the 1990s.
Maybe this is finally the thing that kills it.
But if the last 26 years tell us anything, betting on the next technology trend finally killing Active Directory has been a losing bet.
Long live Active Directory! 🖖
All the best
Spencer