AI has transformed the way procurement lives and breathes.
Over the past few years, the workplace has seen an explosion of new digital tools flooding the market, each offering ways to deliver time and cost savings previously unimaginable a decade ago. Indeed, AI is touching all corners of the procurement function and those who don’t embrace today’s technology are set to be left behind.
One of the biggest drivers in recent times has been the acceleration of generative AI. Since ChatGPT burst onto the market in November 2022, chatbots and large language models have exploded in popularity and surged in use. According to McKinsey’s research, it is estimated that GenAI could add up to $4.4 trillion to the global economy annually while increasing the reach of AI by 15% to 40%.
But two and a half years later, a new kid on the block has emerged – agentic AI. The technology refers to a type of artificial intelligence system that acts as an autonomous agent that is capable of making decisions, taking actions and learning from interactions without the need of human intervention. The key difference from chatbots such as ChatGPT or Copilots is that generative AI is based on data it has learned and is primarily static, whereas agentic AI is continuously processing new information and learning from its environment.
Agentic AI: Operating with minimal human intervention
Nicolas Walden, Europe Practice Leader, Procurement Advisory for The Hackett Group, believes that one of the core advantages of agentic AI is that it is ‘always on’. “Agentic AI, by contrast, is a newer and more autonomous type of AI that can make decisions, take actions, and learn independently,” he tells us. “It can solve more complex problems by analysing data, setting goals, and adapting to new information. Agentic AI is designed to operate with minimal human intervention. It is used to automate workflows and business processes, frequently through a combination of intelligent agents. Examples include autonomous vehicles, cybersecurity threat detection, black-box trading algorithms, and logistics route optimisation and planning. While GenAI capabilities have been widely deployed for around two and a half years, the development of agentic AI is still in its early stages.”
Manoj Chaudhary, the CTO of AI integration company Jitterbit, believes agentic AI is the next frontier of procurement innovation and helps overcome challenging tasks through immediate support to help support actions. “AI serves as a digital partner, accelerating processes and democratising data access. The latest advancements—agentic AI and AI agents—take this partnership further, helping tackle increasingly complex tasks and offering real-time support for decision-making,” he says. “These concepts represent two different approaches to use cases for AI, yet are often grouped together by developers. We see agentic AI as an autonomous decision-making capability that can act independently within defined parameters, allowing enterprises to delegate operational decisions that do not need oversight but still prioritise ethical standards, data integrity, and security. This approach ensures that businesses can achieve efficiency and innovation without compromising on accountability and control.
“AI Agents, on the other hand, are task-oriented for predefined actions and follow specific user commands to automate workflow management, assist in complex development tasks, and provide real-time decision-making support. These agents offer immense potential to improve key areas of business process management with orchestrated oversight, including workflow automation, resource allocation, and performance monitoring.”
GenAI vs Agentic AI: What’s the difference?
While Simon Geale, Executive Vice President at Proxima, explains that generative and agentic AI are likely to be used interchangeably over the coming months, in addition to the real likelihood of other forms of AI being included too. “At their core, the difference is that generative AI creates content based on data (sometimes referred to as large language models, or LLMs), and agentic AI is able to carry out processes and tasks, making decisions based on learned logic, which includes data. You might say that GenAI is more like an assistant who thinks and creates and agentic AI is more like an employee who decides and does. You might also observe that the combination of the two has the potential to be quite special, automating not just processes, but also production.
“In procurement terms, this takes us further down the path of not just full automation, but moreover automated augmentation; we will be able to do more, faster on the current proviso that process and data are in decent shape. That said, whilst generative AI depends on accurate data (the narrower and more precise the better), it can also be part of the solution to getting to accurate data, using a balance of inputs and probabilities ‘to get to clean’.”
AI’s next step
And Burley Kawasaki, VP of Product Marketing and Strategy at Creatio, affirms agentic AI is the next leap forward in enterprise automation – moving beyond thinking to actually doing. “While generative AI produces content like text, code, or imagery by recognising patterns in data, agentic AI goes further and actually takes action. It executes tasks autonomously, moving beyond suggestions to actual orchestration.
“By combining machine learning, automation, and natural language processing, agentic AI can make decisions and manage workflows with or without human input. This shift from output to orchestration is what makes it so impactful. In customer relationship management, that might mean qualifying leads, responding to service requests, or even orchestrating personalised customer journeys, all without manual input.
“Crucially, the real value lies in balancing the AI triad. Predictive AI offers foresight, generative AI brings creativity, and agentic AI drives execution. Together, they create a powerful framework for AI-native automation across the enterprise that drives quantum leaps in intelligent productivity and results.”
Future-proofing procurement transformation
Looking forward, Chaudhary believes both technologies will be increasingly relevant throughout the remainder of 2025 and beyond. “These complementary systems will redefine business operations, setting new standards for productivity, strategic execution, and adaptive management and enterprises that harness both capabilities effectively will be able to carve out a competitive edge in an increasingly automated world.”
As a result of the boom in demand for new technology solutions, there is a fear among some sections of the industry that AI is here to take human jobs. But Kawasaki stresses this is not the end goal. As AI advances, the real opportunity lies in augmentation – freeing human talent from manual tasks to focus on creativity, strategy, and innovation. “When paired with no-code platforms that enable non-technical users to build AI-driven workflows, agentic AI becomes not only powerful but widely accessible. Ultimately, the future of enterprise AI won’t be defined by what it creates, but by what it enables us to achieve.”