EY Principal of Supply Chain & Operations Sumit Dutta says GenAI use cases show promise but that adoption is hampered by skills shortage and legacy
The adoption of GenAI in supply chain is being hampered by cybersecurity issues, a lack of skilled staff and problems with legacy systems, a report from EY says.
Sumit Dutta is Principal of Supply Chain & Operations at EY, and says that although technology can be “transformative” there are “too many roadblocks” hampering its use.
In his paper, ‘How generative AI is used in supply chains’, Dutta says that although GenAI is “gaining a foothold in supply chain management” research from EY shows there is urgent need for system integration, suitable talent and data security if organisations are to enjoy meaningful business benefits.
Dutta’s paper poses three key questions that procurement and supply chain professionals need to ask regarding GenAI:
- What are the key applications of GenAI in supply chain management?
- What are the main challenges supply chain leaders face in adopting GenAI and how can they overcome these?
- What does a successful GenAI programme look like?
Dutta says that in supply chain planning AI has been seen across supply chain control towers, digital twins for virtual modelling and the streamlining of product development.
In sourcing, he points out that GenAI is used across contract management, real-time inventory analysis and vendor risk assessment.
From a wider supply chain perspective, he says GenAI is proving valuable in manufacturing, being used in warehouse automation, predictive scheduling and task management.
“Major retail chains and healthcare industries are already piloting the use of GenAI for tasks such as summarising and analysing customer feedback and generating novel small-molecule entities for drug discovery,” says Dutta.
EY: GenAI roadblocks hampering adoption
He adds: “Although these gains are promising, GenAI has its own roadblocks in terms of data security, talent procurement and incompatibility with legacy systems. Added to this is the cost and complexity of achieving regulatory compliance.
Dutta insists that implementing GenAI into workflows poses far more than mere technical challenges.
One ‘roadblock’ is talent acquisition, says Dutta: “Many organisations struggle to recruit and retain specialised talent for GenAI projects
“From software engineers to business analysts and data scientists, the demand for skilled professionals outstrips the available supply, hindering implementation and development efforts.”
He adds that legacy system compatibility is another problem: “Older systems often lack compatibility with the advanced requirements of GenAI.
“Outdated technology stacks and architectures may lack the necessary interfaces or APIs for seamless integration, necessitating extensive customization and redevelopment efforts.”