A Gartner survey found that most supply chain organizations lack formal AI strategies, relying instead on fragmented pilots that rarely scale. To unlock its full potential, supply chain leaders must redesign operating models around AI capabilities from the ground up through strategic change management.
- Only 17% of AI pilots launched by supply chain organizations scale successfully, while 77% believe their current operating models won't support an AI-driven future
- 59% of supply chain organizations focus on short-term ROI under one year, and 62% concentrate on individual use cases rather than transformational approaches
- 55% of CSCOs are unclear on the ROI of their AI investments despite 67% of digital investments now allocated to AI
- Organizations that properly size their change management efforts are predicted to double their ROI on AI initiatives by 2030 compared to those using legacy methodologies
- Building an AI-native supply chain requires redesigning operating models, conducting end-to-end process mapping, restructuring organizational roles, and creating a unified data layer—not simply bolting AI onto existing systems
A common approach to implementing AI has been to reach for the “low-hanging” fruit – identify the less complex work that can be automated. That has been the consensus among AI users in attendance across numerous educational sessions at supply chain conferences.
But that implies jumping feet first into the technology that drives AI. That could look like adding AI tools or functionality to traditional workflows, but Global research firm Gartner, which guides companies in technology selection, mandates that supply chain leaders take a different tactic: build an AI foundation that starts with change management.
Gartner Director of Research Snigdha Dewal said during a webinar that it’s necessary when building an AI-native supply chain from the ground up. She said chief supply chain officers (CSCOs) have been responding to the market, piloting new use cases, bolting AI onto existing workstreams, and encouraging teams to learn about and adopt AI tools.
“The current operating models and approach to AI shows the lack of a clear strategy and is proving inadequate to set them up for success in that AI-dominated future,” Dewal said.
She noted that CSCOs are being cautious and focusing on short-term return on investment (ROI). According to a Gartner survey, 59% of supply chain organizations are targeting ROI realizable in under one year, and 62% are focusing their efforts on individual use cases, while only 17% are taking a transformational approach to AI.
Even as 67% of supply chain digital investments are now allocated to AI, 55% of CSCOs are unclear on the ROI of their AI investments, according to Gartner.
“Three out of four CSCOs now view AI as essential to driving growth and resilience, yet most admit their supply chain operating models are not built for an AI‑driven future,” Dewal said. “Incremental add‑ons to legacy systems aren’t enough. To unlock AI’s full potential, CSCOs must embrace a bold shift toward the AI‑native supply chain: an operating model intentionally designed around AI’s strengths rather than human constraints.”
The change management approach to AI
To document supply chain AI use cases and clarify the ROI, Gartner surveyed 135 senior supply chain leaders from January through April 2026. Because AI has and continues to evolve so quickly, Gartner notes that use cases are outrunning companies’ abilities to develop the effective change management strategies needed to support it.
Gartner is directing CSCOs to allocate finite change management resources across AI initiatives, instead of toward one-size-fits-all approaches, to achieve broader goals.
“Organizations are getting better at executing change for individual initiatives,” said Lorraine Gavin, senior principal analyst in Gartner's Supply Chain practice. “The bigger challenge today is deciding where to invest limited change management resources so that they support the business outcomes that matter most. AI is making that decision more important than ever.”
Gartner predicts that by 2030, organizations that rightsize their change management efforts will double their ROI on AI initiatives compared to those that continue to rely on legacy methodologies.
Gartner research distinguishes between change methodologies and change strategy.Gartner
These are some tips from Gartner:
- Establish a Formal AI Change Strategy: Institutionalize AI change management as a distinct strategic discipline that supports the organization's AI technology strategy and creates a foundation for intentional, outcome-driven execution.
- Prioritize and Allocate Change Resources Strategically: Treat change management capacity as a scarce resource and concentrate investment on high-value AI initiatives with the greatest potential to drive supply chain outcomes and ROI.
- Adopt a Composable Approach to Change Execution: Define different change approaches based on the scale, speed and organizational context of each initiative, rather than relying on a single standardized methodology.
- Build AI Change Leadership Capabilities: Develop leaders with business acumen, workforce development expertise and risk management capabilities so they can effectively guide AI-enabled transformation.
Fragmented adoption limits scalability
Dewal noted in the webinar that 77% of organizations think their current operating models will not allow them to succeed in an AI-driven era, yet only 23% of those pursuing AI have a formal strategy.
Julia Heyman, a senior principal analyst in Gartner's Supply Chain Strategy practice, said in her roadmap for building an AI foundation that CSCOs should focus on four priorities: expanding their AI expertise, learning from industry leaders, assessing organizational AI readiness and advancing their role in enterprise AI strategy.
She said it’s not about deploying tools.
“It’s about shaping the operating model, upskilling teams and governing data for scale and trust,” she wrote.
Dewal noted that only 17% of AI pilots launched by supply chain organizations were able to scale successfully. That’s why she insists AI needs to be embedded across every function (AI native).
“AI without scalability cannot deliver on its promise,” she said. “… at full scale, AI can ingest real-time signals (and) adjust plans; agents can coordinate across carriers, reroute shipments, resolve disruptions autonomously and interact directly on negotiating in transacting.”
During the Gartner webinar, only 9% of attendees marked in a poll that they were completely familiar with the term AI native, and 40% marked that they were not familiar at all.
Dewal likened it to a prominent shift in the aviation industry: prior to the 1930s when the jet engine was invented, aircraft were made of heavy wooden frames. But the jet engine couldn’t be bolted onto those frames, so aircraft was then built from the ground up around the new engine’s properties and power.
It’s much the same for AI in the supply chain, she said.
“Many supply chains today are still a little bit low maturity, sort of wooden frame,” she added.
Building a new frame
Dewal noted three steps to building an AI supply chain:
- Reimagine the operating model beyond human constraints and around AI’s potential. Old models built, at best, for the digital era limits AI’s capabilities.
- Conduct end-to-end process and core decision mapping. This means dissecting your supply chain into workflows and/or decisions: define who makes the decisions, what data they use, and the current level of automation and maturity of those processes and decisions.
- Use process mapping and generative AI to help create different scenarios of how decisions could be made in an AI-native environment and how that would impact workflows.
- Assess the impact of different levels of autonomy, and identify the most optimal flows; identify workflows no longer necessary in an AI-driven environment. For example, if you automate procurement, do you need purchasing orders?
- Set clear targets for the future state of automation, including the percent of no-touch or automated decisions, and establish the expected accuracy or decision acceptance threshold before full autonomy.
- Redesign the organizational structure with new roles geared toward an AI-native supply chain.
- Assess capabilities and workflows; map new roles and responsibilities from reimagined workflows.
- Create a workforce evolution plan to cultivate collaboration in a human-AI culture in which agents now perform the duties previously done by humans.
- Make targeted investments to restructure and upgrade the tech layer. That doesn’t necessarily mean divesting of an entire existing tech stack. You need a segmented strategy that builds on high-value technologies while selectively developing AI capabilities to help maximize long-term ROI.
- Assess the existing tech stack: What’s good enough? What’s manual or outdated? What’s new, where AI needs to be built from ground up.
- Build a unified data layer independent of functioning silos to sit on top of existing transactional systems. Dewal noted that you can build good AI capabilities without good data, but you need to think about what data you need and why you need it as you change your decision-making structure.
- Build an autonomous orchestration layer that pulls insights from multiple AI agents, forms an integrated point of view, answers questions, gives recommendations and executes actions without continuous human direction.
One of the biggest misconceptions in this process, Dewal said, is that AI-native means rip and replace.
“You don’t have to start from a blank slate. New technology layers can be built on top of your existing tech stack,” she said. “Bolt-on AI efforts will also yield results. I am definitely not denigrating that, but … those gains are likely to remain local optimizations if you don’t start thinking about how to scale your AI usage.”
