New Study Reveals Disconnect Between AI Approval and Usability in the Workplace
Saturday 25 October, 2025 – Artificial Intelligence (AI) has quickly become an essential tool for organizations, with 78% of companies using AI in at least one business function, according to McKinsey’s 2025 State of AI report. However, a recent survey by Cybernews has exposed a significant gap between IT approval and real-world usability of AI tools in the workplace.
While 52% of employers have approved or provided AI tools for their workforce, only 33% of employees using those approved tools say they fully meet their work needs. This disconnect has led to a surge in “shadow AI” usage, with 59% of employees admitting to using unauthorized AI tools. This practice not only exposes companies to potential security breaches but also highlights the failure of governance structures designed to prevent these very vulnerabilities.
According to Žilvinas Girėnas, head of product at nexos.ai, a secure all-in-one AI platform for enterprises, the approval paradox is a result of a flawed procurement and implementation process.
“This isn’t a user problem but a procurement and implementation crisis. We’re approving AI tools on promises and checklists, not on how well they fit work practices. Insufficient tools lead employees to bypass approval, risking customer data on unknown platforms,” he says.
The Rise of “Shadow AI”: A Conflict Between Productivity and Risk
The failure of approved AI tools has given rise to “shadow AI,” where employees use unauthorized software and platforms to get their work done. This practice creates a massive blind spot for IT and security leaders, as 75% of employees admit to sharing potentially sensitive information through these unauthorized tools.
This clash between the company’s need for security and control and the employee’s interest in productivity and efficiency has put leadership in a difficult position. They must either block the use of unapproved tools and risk losing a critical productivity edge or permit their use and lose control over sensitive data. The potential risks of the latter are tangible, with data breaches costing companies an average of $670K, according to IBM.
The root cause of this problem lies in the lack of employee involvement in the procurement and selection process. Many organizations rush to participate in the AI boom and make decisions based on vendor promises or security checklists without fully understanding their teams’ day-to-day workflows. This often results in “innovation theater,” where companies adopt AI tools superficially without integrating them into the business. As a result, employees are left with sanctioned tools that do not solve their problems, leading them to seek out unapproved alternatives.
Practical Steps Forward
According to Girėnas, solving the approval paradox requires a shift away from a top-down, tool-first mindset. He outlines four non-negotiables for organizations to build a secure and productive AI ecosystem:
1. Map employee workflows before selecting tools.
2. Offer a secure sandbox, not just restrictions.
3. Implement a “living” policy with ongoing feedback.
4. Identify internal AI champions.
ABOUT NEXOS.AI
nexos.ai is an all-in-one AI platform that enables organizations to securely deploy, manage, and optimize AI models. Founded in 2024 by Tomas Okmanas and Eimantas Sabaliauskas, who also co-founded several successful ventures, nexos.ai addresses the urgent need for efficient and secure AI adoption in enterprises. Originating in the ecosystem of Lithuania-based tech accelerator Tesonet, the company received €8M in investment in early 2025 from Index Ventures, Creandum, Dig Ventures, and prominent angel investors.
To learn more about nexos.ai and their mission to drive secure AI adoption, visit their website at https://www.nexos.ai/.
Derick is an experienced reporter having held multiple senior roles for large publishers across Europe. Specialist subjects include small business and financial emerging markets.