Follow the AI Footpaths

Source: Towards Data Science

When you walk through a city park, you’ll often see narrow dirt trails cutting across the grass—between sidewalks, across lawns, and through corners that planners never intended people to use.

What Are Desire Paths?

Desire paths are informal routes that form when people choose their own shortcuts instead of the official walkways. Over time, the grass disappears and the trail becomes a visible record of how people actually move through a space.

• Historically: Planners treated these paths as mistakes.
• Today: They’re seen as valuable feedback, showing where a design fails to match human behavior.

The Same Phenomenon Inside Organizations

Employees are already forging their own “desire paths” with artificial intelligence:

• Marketing: Drafting campaign copy with a language model.
• Finance: Summarizing reports using an AI assistant.
• Product: Testing ideas through generative tools.

Often, this experimentation happens quietly, outside official systems or policies—mirroring the informal trails that appear in parks.

Shadow AI: The Emerging Workplace Phenomenon

The term Shadow AI echoes the older concept of shadow IT—when employees installed software without approval from corporate IT departments. Today the pattern repeats itself with artificial intelligence: workers bring generative‑AI tools into their daily workflows long before organizations establish governance structures or approved platforms.

Why It Matters

• Data exposure – Sensitive corporate information can flow to external systems without visibility into how that data is processed or stored.
• Regulatory risk – Frameworks such as GDPR or the EU AI Act may be unintentionally violated.
• Loss of oversight – Security teams struggle to track how information moves through the organization.

Beyond the Risks

Focusing only on risk misses a crucial insight: Shadow AI often reveals where existing systems are no longer keeping pace with how people need to work. Like desire paths in a park, it exposes where employees are searching for faster, more intelligent ways to complete everyday tasks.

How Widespread Is It?

• Adoption – Surveys indicate that nearly four out of five people using AI at work bring their own tools rather than relying on employer‑provided systems.
  • Source: Salesforce Generative AI Snapshot Research
• Personal accounts – Many interact with AI through personal accounts instead of enterprise platforms designed to protect sensitive data.
  • Source: Microsoft Work Trend Index
• Data leakage – More than half of employees admit to entering confidential information into AI systems.
  • Source: ACM article on Shadow AI threats
• Business impact – Organizations experiencing widespread Shadow AI usage report higher breach costs and greater exposure to regulatory risk.
  • Source: IBM Data Breach Report

The Bottom Line

Artificial intelligence is already spreading through workplaces at scale. Governance, training, and security frameworks are arriving later, creating real risks and revealing how technological change actually unfolds inside organizations. Recognizing and addressing Shadow AI is essential for aligning innovation with compliance and security.

Shadow AI as an Organizational Signal

Shadow AI—the use of generative‑AI tools outside official channels—does more than bypass governance. It also exposes the friction points in existing workflows.

Why Shadow AI Matters

A Parallel from Urban Planning

“Cities should be designed around how people actually move through them, not around how planners imagine they should.” – Jane Jacobs

Just as informal footpaths reveal real movement patterns in a city, Shadow AI reveals real work patterns in an organization.

Turning a Threat into an Opportunity

1. Shift the mindset – view Shadow AI as an early indicator of high‑value AI use cases, not merely a governance failure.
2. Map the experiments – collect examples of unofficial AI usage across teams.
3. Identify common pain points – look for repetitive tasks that employees are already trying to accelerate.
4. Prioritise governed solutions – build secure, scalable tools that address the most‑frequent needs.

Benefits of a Curiosity‑Driven Approach

• Risk reduction – fewer unsanctioned tools means a smaller attack surface.
• Employee empowerment – secure tools align with how people already work.
• Productivity boost – automation and augmentation target the tasks that matter most.
• Strategic insight – the “chaos” of fragmented experiments becomes a roadmap for AI adoption.

TL;DR

• Shadow AI is a signal, not just a problem.
• Treat it like an urban planner treats informal footpaths: observe, learn, and redesign.
• By curating the insights from these experiments, organizations can turn AI from a liability into a multiplier of creativity and innovation.

Learning from the AI Footpaths

Organizations that want to govern artificial intelligence effectively must first understand how it is already being used.

1. Gain Visibility

• Shadow AI is a signal, not just a compliance issue. It shows where employees are trying to move faster than existing systems allow.
• Methods to uncover hidden usage:
  • Employee surveys
  • Technical audits
  • Open cross‑department discussions

These activities often reveal that marketing, sales, finance, HR, and product teams are the early adopters.

2. Shift from Suppression to Structure

Once patterns are visible, the focus moves to:

1. Defining which tools are appropriate.
2. Establishing governance policies aligned with data sensitivity and regulation.
3. Designing processes that reflect how work actually happens.

3. Culture Matters Just as Much as Policy

• Employees should feel safe discussing AI experimentation rather than hiding it.
• Fear of punishment or extra workload does not stop experimentation—it pushes it further into the shadows.

4. Enable Responsible Experimentation

5. Turn Shadow Experiments into Coordinated Progress

Understanding what already exists in the shadows is often the first step toward building a resilient and intelligent AI strategy. By making hidden work visible, aligning policies with real workflows, and fostering a supportive culture, organizations can transform scattered experiments into organized, value‑driving initiatives.

A Final Thought

In practice, Shadow AI is rarely the result of malice. More often it reflects misalignment and a lack of communication inside the organization. When employees feel unsafe sharing their experiments, and curiosity is met primarily with correction, the predictable outcome is silence.

People do not stop experimenting.
They simply stop sharing.

If organizations want to govern AI effectively, they must begin by creating environments where thoughtful exploration is possible. Training, practical examples, and clear guardrails make responsible experimentation visible instead of hidden.

Culture Is Key

• Curiosity over suspicion – encourages open sharing.
• Clear guardrails – give teams confidence to experiment safely.
• Visible support – leadership should model and reward responsible innovation.

When curiosity replaces suspicion, experimentation moves out of the shadows and into the open.

The first step toward governing Shadow AI is simple: understand where people are already walking.

About Aleksandra Osipova

• Founder of Apricity Lab, where she helps leaders and organizations navigate the transition to AI‑enabled systems.
• Writer on topics such as artificial intelligence, systems thinking, and the future of work.
• Explore more of her work and insights on her LinkedIn profile.