The Single Biggest Barrier to AI Adoption Isn't the Technology — It's This | Errol Gardner of EY

What It Covers

Errol Gardner, EY's global consulting leader, examines where agentic AI actually stands in enterprise adoption, why human resistance — not technology — blocks deployment, and how large organizations like EY's 400,000-person firm are navigating GenAI and agentic workflows in real production environments.

Key Questions Answered

• •Agentic AI adoption scale: Enterprise agentic AI sits below 1 out of 10 on an adoption scale, with only roughly 20% of organizations using it in any production capacity — and even then, only within a small fraction of their overall business operations. Cloud took 15 years to reach approximately 7 out of 10, and agentic AI faces steeper implementation hurdles.
• •Data governance before experimentation: Organizations must establish private LLMs within their own firewall before encouraging employee AI experimentation. Without this, employees using consumer AI tools will inadvertently expose corporate data to public training models. EY built a controlled internal LLM first, then rolled out access to its 400,000 employees with monitored guardrails.
• •Change management outweighs technology readiness: The primary barrier to enterprise AI deployment is human resistance across all organizational levels — leaders, middle managers, and frontline workers. Workforce anxiety about job displacement drives active resistance to enterprise-grade agentic systems, making structured communication from the C-suite about AI philosophy a prerequisite for adoption.
• •Intergenerational leadership gap: Younger employees readily experiment with AI tools, but frequently report to senior leaders from generations less inclined to adopt or reward AI-driven innovation. Organizations deploying AI should audit whether management layers are actively incentivizing AI experimentation or inadvertently penalizing it through traditional performance and output measurement frameworks.
• •Consulting shifts from inputs to outputs: AI forces consulting firms to move away from billing measured by hours and days worked toward outcome-based delivery models. EY uses this internally — AI enables faster deliverable production, which compresses traditional time-based billing assumptions and requires renegotiating how client engagements are scoped, priced, and measured.