Is rising AI adoption across the US workforce creating shadow AI risks?

AI tools have rapidly become essential to the productivity of American workers, so much so when their company-provided solutions fail to meet their needs, many employees are turning to unauthorized platforms. This phenomenon, known as “shadow AI,” highlights the shortcomings of certain corporate IT investments. It also introduces significant risks, including data leaks, compliance violations and a loss of control over sensitive business information.

But just how prevalent is shadow AI? A recent IBM-sponsored study reveals that while 80% of American office workers use AI in their roles, only 22% rely exclusively on tools provided by their employers. The issue is pronounced among Gen Z employees surveyed (aged 18–24), with 35% reporting they are likely to use only personal AI applications rather than company-approved ones.

In contrast, only 14% of employees in other age groups said they would do the same. These findings make it clear that many organizations are falling short of fully leveraging their enterprise-grade AI investments, leaving workers to seek alternatives that could compromise data security and efficiency.

Workers across the board see AI as a powerful tool that enhances their job performance. Our survey found that a decisive 97% of surveyed office workers in America believe that AI boosts their productivity, with 75% reporting moderate to significant improvements.

Notably, nearly one third of respondents estimate that AI saves them up to six hours each week by helping them complete tasks more quickly and accurately while managing their workloads more effectively. This, in turn, allows them to focus their energy on more creative work.

Still, despite the clear benefits of AI, only 33% of survey participants believe their employers are fully utilizing its potential. Even more concerning, 30% of the youngest demographic identify leadership resistance as the biggest barrier to better AI adoption within their organizations. It’s apparent that a cultural problem exists—one in which employees are making their own decisions about AI use, rather than following organizational blueprints.

To close the gap between employee needs and enterprise AI adoption, companies must strike a balance between ease of use and robust governance.

Nearly 40% of workers say they prefer external AI solutions for their better features. To compete, organizations need to invest in data infrastructure that maximizes their AI tools’ utility. By making internal data cleaner and more accessible to AI models, businesses can begin integrating these platforms into existing systems, delivering productivity and value-creation benefits that consumer-grade platforms simply can’t match.

Finally, businesses must invest in training. 60% of employees say hands-on learning would boost their AI usage. To encourage the shift from shadow AI to enterprise-grade platforms, businesses need to demonstrate more clearly what it’s capable of—along with outlining the shortcomings and risks of unsanctioned AI use.

With 80% of today’s workers expecting AI to play an important role in their work over the next 3–5 years and 50% labeling it as very important or even essential, organizations must take steps now to ensure they’re using it properly. By investing in secure, enterprise-grade AI tools that are intuitive, seamlessly integrated and built on clean, reliable data, businesses can empower employees to excel while mitigating the risks of shadow AI.

In collaboration with Censuswide, IBM surveyed 3,000 full-time office workers who are at least 18-years old, not sole proprietors and are familiar with AI tools. Respondents live and work across North America, evenly split across the US, Canada and Mexico. This blog post is based on the responses of the survey from 1,000 full-time American office workers.

Learn more