IDC, GuidePoint: enterprise security teams outpaced by non-human (AI agent) identity growth
2026-09-15
Per CIO Dive, IDC and GuidePoint Security reported that non-human identities, including AI agents, are growing beyond the ability of existing enterprise systems to track them, leaving security teams outflanked.
Significance 3: A named research-firm finding that enterprise security tooling can't keep pace with agent proliferation is a credible operational/adoption signal, though it comes through a single press summary without the underlying IDC/GuidePoint report in hand.
Implications · machine-drafted, not owner judgment
This is an operational-lane failure mode showing up at scale: organizations are deploying agentic AI faster than they can govern the identities and permissions those agents hold — the AI Mirror thesis's core claim (agentic systems remove the slack that used to hide organizational control gaps) showing up as a security statistic rather than an anecdote. This is direct evidence for the operational-lane argument in Built to Be Replaced and Work That Holds.
- A publicly available IDC or GuidePoint report with the underlying numbers.
- A disclosed breach or incident specifically attributed to an ungoverned non-human/agent identity.
Sources
- press Security teams increasingly outflanked by AI agents retrieved 2026-09-15