How generative AI can drive supply chain transformation

OpenAI’s release of ChatGPT for public users at the end of November 2022 turned out to be an extraordinary event. For the first time, anybody in the world could experiment with what generative AI makes possible.

Open source generative AI and large language models like ChatGPT have been available for several years from companies including Google, Microsoft, Meta, and others. But the democratization of the technology could unlock new benefits, particularly in the world of supply chain.

Expectations related to generative AI are high in the supply chain arena. In a globalized, data-driven business world, AI-powered technology powers efficient and sustainable flow of goods, components, and materials. IDC predicts that in 2026, 55% of the Forbes Global 2000 OEMs will have redesigned service supply chains based on AI. This will ensure that the right parts are available and positioned to solve issues before failure.

Global organizations want more supply chain visibility to tackle issues like cost escalation, and demand volatility, according to IDC surveys. Generative AI has the power to do all that and more, helping businesses improve their transparency, efficiency, and overall resiliency.

Here are a couple of roles that generative AI could play in a supply chain organization:

Data Analyzer: Technology can analyze data from various sources — including purchase orders, invoices, and shipment tracking information — to identify patterns and potential areas for improvement. It can create unique insights through the contextualization of publicly accessible data and information from enterprise resource planning (ERP) systems, warehouse management systems (WMS), and customer relationship management (CRM) systems.
Bottleneck-Free Workflow Enabler: Generative AI, which is quite good at summarization, could be used to review the material/product record and provide a succinct summary relevant to the logistics planner or warehouse operator at the point of use. This enables the reduction of bottlenecks that might hobble process workflows.
Training assistant: Technology can generate process guidelines and training material for internal and external use (e.g., supplier management guidelines, including delivery conditions updates or sustainability frameworks). For training manuals and updates, generative AI could identify best practices and create a comprehensive document. However, training materials/guidance must still be validated and approved by the relevant person/product owner (i.e., there must be a defined approval and editing process).
Sustainability Tracker: Technology can assist the creation of sustainability reports by providing data analysis and insights into performance. A new category of report could be created (and its production automated) if the outputs of generative AI-powered technology and intelligent automation technology are combined.
Provider of Mutual Understanding: In any multinational/multicultural environment, the way people communicate and share information really matters. Generative AI technology is not just an AI-powered translator — it can enhance text with understandable, information-based insights. This can accelerate information transfer and significantly reduce problems of being “lost in translation.” The result may be improved trust and collaboration, enabling more efficient and effective decision making.

Embedding generative AI into ERP, SCM, supply chain control and planning tools will become standard — but this is not the limit. By combining generative AI with low-code/no-code development tools, for example, deep technical expertise would not be necessary to build new custom applications powered by generative AI.

But to obtain the full benefits of generative AI — or any other new technology, for that matter — organizations must test, develop, deploy, and create focus groups that study the application across processes. Tools like ChatGPT democratize AI in daily operations, though using it for business purposes requires being 100% sure about its accuracy, reliability, and scalability.

Users and developers also cannot forget security. Make sure the technology you are using includes safeguards against bias and “jailbreaking,” or tricking AI chatbots into disregarding filters intended to prevent the generation of dangerous or hateful content.