Generative AI is now widely used by consumers for everything from planning vacations to summarizing the day’s news to organizing shopping lists. The stakes are low for everyday users: A flawed AI-generated restaurant recommendation doesn’t amount to much more than an inconvenience.

However, this growing comfort with AI is making its way into the workplace as well, with the share of U.S. employees using AI at work increasing from 21% to 40% in just two years. Employees and the companies they work for are increasingly seeking ways to leverage AI to boost productivity and efficiency. 

Yet, when it comes to the supply chain, particularly risk management, AI adoption looks very different. Usage remains uneven, cautious and often invisible. Despite 94% of supply chain leaders planning to use AI in the next two years, only 7% have fully scaled it across their operations. 

The difference is understandable when it comes to supply chain risk management. Decisions based on AI carry more weight. Supply chain risk management teams are responsible for lowering incident rates, identifying hazards, strengthening safety programs, and meeting regulatory requirements. Here, mistakes can have more dire consequences: supplier disruptions, compliance failures, reputational risk or operational setbacks. These risks extend far beyond inconvenience, which helps explain why adoption across the supply chain has progressed more slowly than some leaders expected.

Supply Chain Managers Need to Approach AI Differently

Supply chain risk managers operate in an environment where accuracy, accountability, and documentation are critical. Decisions often need to be audited and justified to internal stakeholders, regulators, customers and suppliers alike. In environments where confidence is critical, unvetted AI tools can introduce too much uncertainty. The hesitation is a natural response to unclear governance and undefined expectations.

This caution can also explain why AI adoption remains uneven despite strong executive enthusiasm. Executives and leadership teams may view AI as the tool needed to transform and modernize their supply chain operations, while procurement and risk management teams are still working to understand the technology, establish guardrails, and define responsible-use policies before deploying it at scale. The result is a growing disconnect between leadership assumptions and operational reality.

However, ignoring AI altogether means missing opportunities. AI has the potential to reduce administrative burden and accelerate decision-making. It can also improve visibility across increasingly complex supplier networks, ultimately accelerating business processes and boosting ROI. All of these benefits can free procurement and supply chain risk managers from repetitive tasks, allowing them to focus on what matters most.

But to secure these benefits, they first need confidence in the tools’ outputs.

The Visibility Gap: AI Use Is Happening Quietly

Without clear expectations and guidance from leadership, AI adoption often moves underground. For some teams, this means avoiding AI entirely and continuing business as usual, even if that results in missed opportunities.

For others, it means experimenting privately, quietly summarizing audit findings, analyzing contractor data, or drafting communications without integrating those practices into shared workflows. Over time, this creates shadow AI sprawl, the uncontrolled, unmonitored or unauthorized use of AI within an organization.

While this silent experimentation may help employees become more comfortable with AI, it also creates a new layer of risk. When employees use unsanctioned or inconsistent AI tools without proper operational enablement, expected efficiency gains stall, workflows remain fragmented, best practices fail to scale, governance weakens, and proprietary company information is divulged. In supply chain risk management, where auditability and process consistency are critical, disconnected AI adoption can make it harder to validate outputs, maintain documentation standards and ensure accountability.

These disjointed AI adoption patterns help explain why promised AI-driven efficiency gains have often materialized more slowly. Leaders may believe they have an AI strategy in place when, in reality, they have isolated pockets of experimentation.

The Real Challenge: Building Trust

The challenge in deploying AI at scale is increasingly organizational and people-related. Procurement and supply chain risk management teams often lack the confidence, enablement, and psychological safety needed to use AI effectively and responsibly. Without these components, employees remain hesitant, teams work inconsistently, and organizations struggle to scale adoption.

At the heart of the matter is trust. Procurement and supply chain risk managers need to trust not only the AI tools they are using, but also the organizational structure around them.  

This trust gap most often reflects operational uncertainty. When individuals understand expectations, feel supported, and know how to use these tools, they are more likely to use them with confidence. Conversely, when employees lack clarity around governance, accountability and acceptable use cases, even the most capable AI tools can lead to risky decisions.

The good news is that trust can be built. Leadership modeling, practical training, transparency in governance, and workflow clarity enable trust, allowing all involved to reap the benefits of this new wave of technology. For procurement organizations, those that benefit the most from AI may not be the earliest adopters, but those that operationalize trust most effectively.

Defining the Next Wave of AI Adoption Across the Supply Chain

When AI tools are deployed with clear expectations and the belief that responsible use will be supported rather than penalized, AI can be a valuable enablement tool for those in procurement and supply chain risk management.

Supplier networks are becoming increasingly complex, and regulatory pressures are rising. As a result, operational readiness is becoming the defining factor in successful AI adoption. But readiness is an ongoing effort to build confidence, shared standards and clear expectations across teams. Simple but proven rollout strategies can do the trick:

Establish clear policies. Straightforward policies with understandable dos and don’ts eliminate confusion and allow employees to experiment responsibly in the open.

Train employees. Showing employees how to use new tools builds their confidence and maximizes their effectiveness.

Prepare for some resistance. Adoption takes time. Keep an open line of communication and gather feedback on where the tools are working and where adjustments are needed.

Plan, do, check, act. As with other evolutions in operational workflows, an open feedback loop is essential to implementing AI. After the planning phase and the initial rollout of AI tools, leaders should take the time to check in with their employees and act on any changes that need to be made, be those updates to the responsible use guidelines or adding additional training.

The organizations that normalize learning, create shared standards, and make experimentation safe will be the ones that succeed in using AI to improve their supply chains. But success will come down to who implements it most effectively while supporting employees, building trust, and creating the conditions needed to maintain progress.   

Geoff Goodman is principal of solution delivery and change management at Avetta.