Employees are using AI where they know they shouldn’t

Despite widespread anticipation about AI’s positive impact on workforce productivity, most employees feel they were overpromised on its potential, according to GoTo. In fact, 62% believe AI has been significantly overhyped.

However, this is likely because employees aren’t making the most of what these tools have to offer. 86% admit they’re not using AI tools to their full potential, and 82% say they aren’t very familiar with how AI can be used practically in their day-to-day work.

AI could free up valuable time at work

All told, employees estimate that they’re spending 13 hours per week on tasks that could be handled by AI. This means that in the U.S. alone, businesses could be missing out on more than $2.9 trillion annually in greater efficiency.

“Employees are already using AI and are seeing productivity gains, yet despite these benefits, our latest research shows people still view AI as overhyped. While many recognize its value, they don’t yet see it as the revolutionary change they were promised. This gap likely exists because many workers admit they aren’t realizing AI’s full potential or don’t know how to apply it in practical ways,” said Rich Veldran, CEO of GoTo.

Employees lack confidence in AI outputs

Instead of using AI to save time in their day-to-day work, 54% of employees admit to using it for sensitive or high-stakes tasks, such as those requiring emotional intelligence, affecting safety, or involving ethical or personnel decisions, despite knowing they shouldn’t.

86% of employees aren’t very confident in the accuracy and reliability of AI tools, and 76% say they often provide outputs that need to be refined or revised by users.

At the smallest companies, those with 50 employees or fewer, only 59% of workers use AI, and 46% of SMB employees say they don’t know how to use AI to save time or improve their work. In contrast, at larger organizations, nearly 80% are using AI.

“Contrary to what you might think, it’s not just older workers who are struggling to realize the benefits of AI tools,” said Dan Schawbel, Managing Partner, Workplace Intelligence. “Younger workers also admit they’re not using these tools to their full potential. In fact, 74% of Gen Z employees say they aren’t very familiar with how to use AI practically in their day-to-day work. This highlights the importance of equipping all generations with the tools and education to use AI safely and effectively.”

How to close the AI adoption gap

Give employees the tools they want: Employees say an AI virtual assistant, AI tools that automate certain work tasks, AI communication tools, GenAI tools, and an AI chat/messaging assistant to communicate with customers, would be most valuable for them, but only 4 out of 10 say their company offers these.

Improve policies and training to prevent AI misuse: Just 45% of IT leaders say their company has an AI policy in place. Both employees (81%) and IT leaders (71%) believe AI tools need better instructions and guardrails for proper usage. 87% of employees also feel most workers are not being trained properly to use AI tools.

Be purposeful about AI implementation and ROI measurement: At companies using AI, 21% of IT leaders admit their company is adopting AI or buying AI tools just because they think they should, not after careful consideration or with a clear plan in mind. What’s more, 49% of IT leaders say their company isn’t measuring the ROI of AI tools very well.

Recognize that a small investment can have a major impact: 77% of IT leaders say their company would only need to spend an extra $20/month or less per employee on AI tools to save each employee an additional one hour a day in greater efficiency.

Help IT leaders understand the employee perspective: IT leaders and employees aren’t always seeing eye-to-eye when it comes to AI use, practicality, reliability, and more. Companies that take steps to address these disconnects will be well-positioned to maximize the benefits of AI for their organization.

“The reality that many employees are using AI tools for high-stakes or sensitive decisions—often despite internal policies or ethical concerns—raises serious questions about the effectiveness of current AI governance. If misuse occurs without regret or consequence, it signals that policies alone are insufficient unless paired with comprehensive, ongoing education, clear, enforceable standards,” Jesse Booth, Senior Director of Security Operations, GoTo, told Help Net Security.

“For CISOs and security leaders, this presents a clear mandate. Prioritizing AI policy development cannot be separated from the training and cultural change efforts necessary to embed responsible AI use across the workforce. Policies should be living documents that evolve alongside AI capabilities and organizational needs, emphasizing clarity, transparency, and user empowerment. Meanwhile, training programs must go beyond technical “how-to” sessions to include ethical considerations, risk awareness, and scenario-based learning that prepares employees to navigate the nuances of AI decision-making. Risk averse companies should also consider technical controls that can constrain or monitor how employees are using AI tools, to limit chances for misuse,” Booth concluded.