Every business function is making bold claims for the transformative potential of generative artificial intelligence (GenAI). Supply chain is no exception. Indeed, IBM has described GenAI as a “game changer for the supply chain.” Likewise, Deloitte says the technology can “revolutionize supply chain management.”
Given that relatively few organizations seem to have made significant progress to date, such predictions feel a little overblown. Right now, Gartner suggests, fewer than one in six supply chain leaders have begun implementing GenAI solutions (even if more plan to do so).
There is a good reason for this: organizations that have begun trialing GenAI solutions to supply chain challenges report only modest success. The evidence is anecdotal, but it is fair to say that supply chain leaders are underwhelmed by their early forays into the technology.
This should be neither surprising nor discouraging. While GenAI has evolved at breakneck speed over the past two years, identifying use cases for new technologies and perfecting deployment takes time.
The history of less revolutionary forms of AI in the supply chain space underlines this point. Supply chain leaders did not embrace these tools overnight, but now widely regard them as essential. AI tools are good at forecasting demand, for example, supporting precise inventory management. Predictive maintenance is also proving valuable, significantly cutting production downtime.
AI has, therefore, gained traction at an accelerating rate. Research suggests adoption rates will more than triple over the three years to 2025. In light of proven benefits, supply chain leaders have invested heavily in scenario-planning tools and optimization technologies. The ROI is encouraging, although currently far from “game-changing” or “revolutionary.”
“Excel is the supply chain industry’s dirty little secret: IMD’s research suggests supply chain professionals spend as much as 60% of their time on Excel spreadsheets.”
Where, then, might GenAI offer the greatest impact as it evolves? Three areas stand out.
GenAI will prove a cheaper, more efficient replacement for some of the narrower AI tools. Tasks where narrow AI currently predominates, such as calculating optimal inventory levels, are often much less straightforward than they initially appear. For example, the classic news vendor challenge – how many newspapers to buy, given the inherently limited shelf life of the product – requires multiple inputs to reach a reliable answer.
Narrow AI solutions address this difficulty using complex algorithms that are run repeatedly until they produce the hoped-for results. This requires a substantial commitment of resources. GenAI tools are now matching the results achieved by algorithmic models and, in some cases, exceeding them. Critically, moreover, they do not demand the same level of resourcing, representing a much less costly route to inventory optimization.
Excel is the supply chain industry’s dirty little secret: IMD’s research suggests supply chain professionals spend as much as 60% of their time on Excel spreadsheets. Not only does Excel hoover up huge amounts of time in the supply chain function, but it also creates undesirable dependencies. With few agreed standards on spreadsheet creation, individual approaches proliferate. If the individual in question leaves the business, faced with the idiosyncrasies of their handiwork, their successor is often obliged to start from scratch.
GenAI won’t replace Excel, but it does have the potential to take over much of the heavy lifting in spreadsheet development and analysis. That’s because tools such as Microsoft Copilot are now capable of programming Excel on the user’s behalf. This leaves scope for significant productivity gains, as the time devoted to Excel falls sharply and makes the loss of an individual spreadsheet author less of a disaster.
Outside established AI use cases such as demand forecasting, the supply chain function rarely asks other critical questions of the technology because, until now, it hasn’t been able to answer them. During the COVID-19 pandemic, when supply chain disruption left many businesses with limited amounts of raw materials, leaders found it very difficult to produce revised production schedules. All their forecasting tools were set up to answer questions about what inputs would be required for a given output; when the question was reversed, they had nothing.
GenAI, in contrast, can make a stronger attempt. Its flexibility and agility, in contrast with the narrow parameters of more traditional AI, enable supply chain leaders to deploy it in new areas, and it can offer meaningful answers to a wider range of questions.
It’s worth acknowledging that GenAI won’t necessarily provide the right answers, but it does give supply chain leaders access to a larger store of intelligence, which they can then make available to the rest of the business.
In each of these three cases, GenAI’s functionalities will continue to improve and expand, and more examples of areas where it can make a material difference will emerge.
There is no denying the speed at which this technology, as well as the human capability to use it, is moving. Some organizations report that, while GenAI pilots and tests they ran six months ago failed to convince, the same experiments are now generating encouraging results.
The potential gains are considerable. Looking at the Excel example alone, a modest 10% reduction in the workload required for a supply chain function would unlock significant resources. Freed from spreadsheet drudgery, these supply chain professionals can be redeployed to more value-added tasks.
But, realistically, predictions of a revolution look premature. In the short to medium term, GenAI technologies will continue to operate alongside established tools, including narrower forms of AI.
Indeed, in certain areas, many organizations will not feel ready to rely on AI, even if GenAI is beginning to prove to be a superior technology. Work with an ethical dimension, say, or connected to health and safety, will continue to demand human input. The dangers of getting it wrong are too great.
Nevertheless, GenAI solutions are at last becoming viable propositions for the supply chain function. Take-up rates will continue to increase. The relevant question seems to be not if, but when GenAI will be in a position to take over the supply chain.
IMD’s Simon J Evenett, Johannes Fritz, and Tommaso Giardini argue that the Trump administration will take a more sophisticated approach to AI regulation than some expect ...
AI is rewriting the rules of hiring – job seekers and recruiters alike must adapt to a process driven by algorithms but grounded in human connections. ...
Breakthroughs in agentic AI, open-source development, and closer collaboration between AI companies and the defense sector have significantly raised the AI risk profile. These developments have compelled us to advance the AI...
GenAI can deliver substantial benefits, but few organizations understand the risks. They need a deliberate data governance strategy....
Explore first person business intelligence from top minds curated for a global executive audience