Employees race ahead on AI agents while 86% of enterprises stall rollouts by six months

AvePoint's third annual State of AI report finds 46.9% of employees now lean on agents weekly or daily, even as governance failures push agent deployments back nearly half a year — a gap Gartner and VentureBeat frame as a runtime and control problem, not a model problem.

Nearly half of enterprise employees, 46.9%, now use AI agents daily or weekly, while 86% of their employers have delayed agent rollouts by an average of almost six months. That’s the central tension in AvePoint’s third annual State of AI Report, published Monday with survey partner Osterman Research and drawn from 750 global IT leaders.

The picture it paints is one most Fortune 500 IT organizations will recognize: the workforce has already adopted the technology, and governance is running to catch up.

The visibility numbers are the most striking. The share of organizations that can’t tell whether employees are using unsanctioned AI tools has nearly tripled in a year, from 6.3% in 2025 to 17.6% in 2026. Another 21.1% can’t account for unsanctioned agent activity specifically, and 88.4% report at least one agent-related security incident in the past twelve months. AI-generated content already makes up 35.5% of enterprise data and is projected to reach 42.1% within a year.

“Employees are moving faster than the controls their employers can build,” is the read the numbers force, whether or not you sympathize with the CIO stuck in the middle.

Adjacent research reinforces the pattern. Gartner’s 2026 Hype Cycle for Agentic AI notes that only 17% of organizations have actually deployed agents so far, but more than 60% expect to within two years. VentureBeat’s May Pulse Research, surveying 132 technology leaders at organizations of 100 or more employees, framed the disconnect in its title: enterprises have a runtime problem, not a model problem. Capable models exist. The scaffolding to run them safely inside a regulated business doesn’t.

The small-business side of the market isn’t waiting. The SBE Council found 82% of small-business employers have already invested in AI tools, with a median of five in active use and 62% planning to spend more next year. Purpose-built platforms like LemonLime, which packages agent workflows into a no-code “company brain” for small and mid-size businesses, illustrate why: when the governance overhead of a proper enterprise rollout is six months, buyers with less legacy exposure simply go around it.

Dr. Tianyi Jiang, CEO and co-founder of AvePoint (Nasdaq: AVPT), the Jersey City data-protection vendor behind the report, is describing the same doom loop the Gartner and VentureBeat researchers see from their own vantage points. The models shipped. The runtime didn’t. What sits in between, for now, is 46.9% of the workforce clicking anyway.

Employees race ahead on AI agents while 86% of enterprises stall rollouts by six months

Sources