Why Nearly Half of Employees Use Unauthorized AI Tools — And What Leaders Must Do About It

Nearly half of all professionals (45%) have used unsanctioned AI tools at work in the past 30 days, according to WalkMe's State of Digital Adoption 2026 report, based on a global survey of 3,750 professionals. More alarmingly, 36% used confidential business data alongside those tools.

The root cause is rarely malicious. WalkMe CIO Tal Carmi told ZDNET that shadow AI usually signals an adoption problem: employees bypass approved tools when they're hard to use, don't fit existing workflows, or simply underperform. Adding to the problem, 34% of professionals didn't know which AI tools their employer approved, and only 21% had been warned about AI policies.

The stakes are real. Unsanctioned AI use can expose sensitive data and create compliance violations. Under the EU AI Act's Article 50, businesses face fines of up to €15 million or 3% of annual worldwide turnover for non-compliance.

Key Takeaways

• Let employees explore, but track it. Thomson Reuters COO Kirsty Roth recommends monitoring AI use early rather than blocking it. Her team watched where employees went, then worked with them to migrate to secure, sanctioned alternatives.
• Build sanctioned tools that actually work. Chase CIO Gill Haus embedded AI controls into LLM Suite, the bank's internal agentic platform, so staff can experiment safely without breaking rules.
• Education closes the gap. WalkMe's Carmi notes that most unsanctioned use stops once employees understand the risks and are given better approved alternatives.

The consensus from top CIOs: rigid restriction backfires. The winning strategy pairs employee curiosity with clear guardrails and tools that deliver real value inside secure environments.

Read the full article on ZDNET