How can we turn supply chain volatility into foresight? We are at an inflection point for AI, writes Jonathan Jackman, Kinaxis‘ VP EMEA, who discusses impacts on the future supply chain.
In today’s world, warfare, sanctions and climate instability are fracturing global supply chains and upending business plans with little warning. In fact, we must accept that volatility has shifted from being an exception to a defining feature of the operating environment.
In response, organisations are accelerating their adoption of AI, drawn by its promise to improve decision making and build resilience in an increasingly unstable world. Yet, as the enthusiasm for AI grows, so do the risks associated with how it’s being deployed.
Many businesses have already embraced early generative AI tools that operate alongside existing processes, though without fully embedding them. While these systems can speed up analysis, they often lack access to critical data and an understanding of wider business context, resulting in new forms of risk rather than increased protection.
Unlike earlier tools, agentic AI can not only analyse information but simultaneously take action, considerably expanding its potential impact. It also increases the consequences of getting it wrong, though.
When AI systems operate without full situational awareness or clear governance, the outcomes can be immediate and damaging, ranging from misdirected inventory and excess production to costly compliance failures.
This is a pivotal moment for AI adoption; agentic AI will play a central role in the future of supply chain decision making, but its success will depend less on the speed of adoption and more on how responsibly these systems are integrated in core processes.
A choice for leaders
As organisations begin to use AI to help them navigate disruption, they face a clear choice. On one side, generative AI tools and copilots are added onto existing processes, offering quick wins and impressive demonstrations. Yet because they sit outside of the workflows where decisions are made, they rely on fragmented data and produce outputs that lack context and accountability.
In complex supply chain environments, any shortcomings can escalate rapidly, with misaligned decisions leading to undermined trust and increased risk exposure.
On the opposite side, organisations can begin embedding intelligence directly into decision making workflows. At its most advanced, this involves agentic AI systems that operate on real-time data alongside the wider business context, allowing them to coordinate responses across the organisation.
When AI is embedded like this, organisations can move beyond reactive responses and gain the ability to anticipate disruption and act decisively before any issues can escalate.
Designed for human-in-the-loop
With all this, maintaining human oversight and accountability when using AI systems should remain a design requirement. While there are concerns that AI might replace people, agentic systems will only deliver the most value when they are designed to work alongside humans.
People are and will remain responsible for the most important decisions. They define objectives, approve actions with significant impact and remain accountable for outcomes.
Within these outlines, autonomous agents can monitor signals, coordinate activity across functions and generate response options. As a result, human decision makers can then focus on areas where judgement and morals, as well as regulatory understanding, are crucial.
More importantly, embedding agentic AI into decision workflows enables oversight to be applied from the beginning. Unsafe or non-compliant actions can be prevented automatically, rather than identified after the fact. As regulators, particularly in the EU, place greater emphasis on transparency and explainability, this level of control is becoming increasingly necessary.
Trust as the foundation
Supply chains are at risk due to a lack of systems that enable transparent, coordinated decision-making.
As uncertainty and instability continue to rise, advantage will come from adopting AI responsibly and embedding it into core decision processes with clear governance and human accountability.
Ultimately, trust is not the result of faster decisions. It is what makes them possible.
