Artificial intelligence is rapidly becoming embedded in daily work, but employee adoption is outpacing corporate oversight, raising concerns about risk and governance, according to new research.
The survey from Barndoor.ai found that 91 per cent of enterprise employees are now using AI on the job, with many going beyond basic tasks and using the technology to take direct action within business systems. More than a quarter of respondents (28 per cent) said they are already using agentic AI tools that can perform multi‑step tasks on their behalf, while 60 per cent use AI daily or more, including 35 per cent who use it multiple times a day.
“People want AI to do tasks across the apps they use every day, but without proper controls, that’s risky,” said Oren Michels, co‑founder and chief executive officer of Barndoor.ai. “The data suggests a widening gap between how fast workers are moving and how prepared companies are to support them safely.”
The report suggested that employees are increasingly granting AI tools access to core business systems. Half of respondents said they have connected AI to work‑related applications, including email and calendar platforms, as well as customer data, financial systems, internal databases and application programming interfaces.
At the same time, nearly half of employees (48 per cent) reported using non‑approved AI tools at work. Among those, 42 per cent said unauthorized tools deliver better results, while 40 per cent cited pressure to complete work quickly. Nearly one in five employees said they are unsure of their organization’s AI policy.
Use cases have also expanded beyond content generation into operational workflows. Employees reported using AI to automate invoicing processes, summarize customer support tickets, manage software builds, generate reports and analyze sales pipelines — tasks that previously required manual effort or multiple systems to complete.
The findings also point to a lack of standardization in AI use. More than half of workers said they use two to three different AI tools, while a smaller share reported using four or more. Among platforms, ChatGPT and Microsoft Copilot lead adoption, followed by tools such as Gemini, Claude and Perplexity.
Despite widespread adoption, support from employers appears limited. Two in five (42 per cent) of employees said they do not feel well supported in using AI at work, and fewer than half (46 per cent) said their organization has a defined AI policy in place. Among those using company‑approved tools, ease of use, capability and data privacy were cited as the main factors influencing adoption, rather than formal security concerns.
Barndoor.ai said the findings highlight the growing need for clearer policies, stronger governance and better support systems as AI continues to expand across enterprise environments. Without those guardrails, the gap between employee behaviour and organizational readiness is expected to widen further, it warned.
Image credit: Depositphotos.com
