What Shadow AI Tells You About Your Business

This year, the percentage of people who use AI for work shot to 81 percent, from 56 percent in 2024. But is this usage authorized by the organization? 

In reality, one in three employees say their company doesn’t provide the right AI tools, and half would rather turn to unauthorized solutions to get work done. The use of “Shadow AI,” or when someone uses AI without the organization’s knowledge or in direct opposition to guidelines, poses several threats to an organization. From inadvertently exposing sensitive data to bumping into compliance issues, shadow AI puts employers at risk and harms an organization’s reputation. 

Instead of revisiting age-old topics like building a strong AI policy with clear guidelines and training, I want to instead focus on how IT teams can leverage shadow AI to their benefit. Business leaders and IT teams can use the insights and indicators from shadow AI to uncover valuable use cases, understand unmet needs, and build a culture that embraces change.

Why employees use unauthorized AI tools

Generative AI doesn’t just live in a standalone chatbot anymore. It’s proliferated into our software products, apps, browser extensions, and even the browser itself. And with free trials and versions available with just an email address, it’s not difficult to understand why employees are tempted with unauthorized tools. 

At this year’s Microsoft Ignite conference, VP of AI Platform at Workday Dean Arnold says, “Shadow AI happens when we don’t provide employees the adequate tooling they need to do their work day to day. They want to be effective, they want to get the job done, so what do they do? They pull out their personal AI and available tools and use them at work.”

Shadow AI is certainly a risk. But it’s also a clear sign to business leaders that workers crave more support, automation, and capabilities in their day-to-day workflows. Employees want to do good work and they want to focus on the work that energizes them—but they also fear reprisal or getting in trouble at work. And, some worry that their colleagues will think they’re lazy or a fraud if it’s revealed they use generative AI. All these symptoms point to one common denominator:

To combat shadow AI, workplaces need to embrace a culture of open communication, collaboration, and empowerment when it comes to AI tools. A cultural shift can directly reduce risk and support governance while supporting workers where they need it most.

Three actionable methods to channel AI opportunities

Policies, procedures, processes, and guidance—while vitally important to an organization’s IT and cybersecurity structure—can only go so far to disrupt shadow AI. Here are three actionable ways to tackle shadow AI with a strategy that embraces alignment, communication, and engagement:

• Run an (anonymous) survey about what’s in use. Before you can implement any larger initiatives or discovery sessions, you have to know your baseline. Run a truly anonymous survey to learn more about what AI tools employees use, whether they are officially sanctioned by the organization or not. Ask what their use cases are, how the tool helps them work better, and if they think it would benefit others at the organization. Not only does this give you a solid foundation of what shadow AI might exist in your company, but it also gives your employees a safe space to show you where their true roadblocks and pain points are in their workflows. And, they can share information without fear of retaliation from management or judgement from their peers.

• Establish an AI committee focused on use cases. You’ve probably heard the advice of establishing an AI committee to develop your strategic vision and oversight for AI initiatives a dozen times by now. While those activities are important, in this case I encourage an AI committee to address the results of the anonymous survey. Identify consistencies in the survey responses around perceived unmet needs and potential use cases to bring a more governed and secure AI solution to the table. Or, the committee can identify how to resolve these unmet needs with solutions and capabilities already in your tech stack (there are more options than you might think).

• Set regular check-ins. Establish recurring times both within individual teams and with cross-functional groups to discuss workflows, processes, and pain points—both with an AI lens and without one. The most creative and successful AI solutions flourish when they focus on solving real, true pain points across an organization. By fully understanding what doesn’t work well (and what does) in an organization, the team can think more holistically about how AI tools may alleviate problems. 

Shadow AI happens because workers want to move fast and they want to take advantage of all their tech stack has to offer. When policy and procedure only takes you so far, it’s critical for IT teams to move from being technology blockers to enablers in order to embrace all that AI has to offer. A combination of policy and cultural empowerment provides business leaders a clearer picture of AI demand to build a faster path to governed solutions. 

To learn more about how to develop cultural shifts, or to even test the cultural readiness of your organization, reach out about our free 90-minute workshop AI & Appetizers that gives business leaders AI clarity separated from AI hype.