Did you miss Ardent Partners’ recent webinar, Evolution of AI in Procurement: The Agentic Age, that shed light on how artificial Intelligence (AI) has come a long way — from basic rule-based systems to today’s advanced models that can learn, reason, solve problems, and even create.
The session featured industry experts from Ardent Partners, as well as Fairmarkit’s CEO, Kevin Frechette, and Coca-Cola Europacific Partners’ Director – Procurement CoE, Anthony Breach, who explored artificial intelligence in procurement and its evolution from generative AI to Agentic AI that actively pursue objectives and take meaningful action in the real world.
In this four-part series, we feature articles based on the webcast discussion as well as this link to the full event.
Today in Part Three, we explore how Agentic AI marks a major evolution in artificial intelligence by enabling systems to act autonomously toward defined goals, making decisions and executing tasks, like procurement and logistics, without direct human input. Unlike traditional AI models, it combines technologies, such as generative AI, machine learning, and rule-based systems, to create intelligent agents that integrate seamlessly into business operations and user experiences. While offering transformative potential, agentic AI also demands careful planning, ethical oversight, and strategic evaluation to ensure responsible deployment and long-term value.
Agentic AI: The Next Evolution in Intelligent Automation
The conversation around artificial intelligence (AI) is evolving rapidly, and one of the most exciting developments currently making waves is agentic AI. Often referred to as the next step beyond generative AI, agentic AI represents a significant leap forward in autonomy, decision-making, and operational efficiency.
Unlike traditional AI systems, including even the most advanced generative AI models, agentic AI is designed to act independently. Rather than waiting for human prompts or inputs, agentic AI systems can pursue defined goals autonomously, making real-time decisions based on the conditions they encounter. For example, in a procurement context, an agentic AI system could be tasked with maintaining inventory levels. Once inventory drops below a certain threshold, the system could automatically source new suppliers, place orders, and manage the logistics — all without direct human intervention.
New AI Evolution Combines the Best Technologies
This evolution isn’t about replacing existing AI models but rather combining the best of current technologies. Agentic AI leverages machine learning, natural language processing (NLP), generative AI for communication, reinforcement learning, rule-based systems, and evolutionary algorithms. It’s not a standalone AI “type” but a comprehensive approach that pulls from across the AI ecosystem to create intelligent agents capable of managing tasks and solving problems with minimal oversight.
However, as with any breakthrough technology, it’s important to acknowledge the nuance. Terms like AI agents, agentic AI, and autonomous agents are sometimes used interchangeably, but the real power lies in how these systems are trained and deployed. Agentic AI isn’t about checking a box to say, “We’ve implemented AI.” It’s about clearly defining the job to be done, designing the agent to handle that job, and ensuring appropriate safeguards are in place, including monitoring, auditing, and adapting the system over time.
In practice, the degree of autonomy given to an AI agent can vary. More autonomy often leads to increased adaptability, but can also introduce unpredictability. Less autonomy makes an agent’s behavior easier to predict and control, but may limit its effectiveness in complex, real-world scenarios. Early adoption will likely include many human-approved checkpoints, and as the systems mature, those checkpoints may decrease, allowing agents to operate with greater freedom.
Beyond the technical excitement, agentic AI also offers a transformative business advantage: simplifying the adoption of technology. Unlike traditional platforms where users had to learn and interact directly with complex interfaces, agentic AI systems will often sit invisibly in the background, providing outcomes via simple, natural-language interactions. This not only enhances the user experience but also drives faster adoption across organizations.
Shift in Approaching AI
Of course, adopting any advanced AI solution comes with important strategic questions. Businesses must evaluate whether deploying an agentic AI system is desirable and whether it can truly deliver value. They must assess scalability, the capabilities of their existing teams, and the anticipated adoption curve. Crucially, the ethical and compliance frameworks must be in place to ensure AI behavior aligns with enterprise policies, industry regulations, and societal expectations.
Many organizations are now approaching this shift with structured, strategic frameworks. They are balancing innovation with governance, working closely with trusted technology vendors, and building partnerships with startups to explore new capabilities in controlled environments. Interestingly, procurement teams, traditionally seen as operational, are emerging as drivers of AI innovation by identifying practical applications and fostering collaboration across functions.
Another important shift agentic AI brings is the relationship between IT and business units. Historically, IT teams played a central role in system implementation, particularly during the ERP era when complex integrations were required. The advent of Software-as-a-Service (SaaS) shifted some of that control back to business functions. Now, with modern AI — especially agentic AI — the collaboration is evolving again. Pre-tested APIs and intelligent platforms reduce the technical burden, allowing IT to focus on oversight and architecture rather than deep system integration.
One final but critical consideration is cost. As AI-powered services mature, pricing models are still in flux. Organizations need to pay close attention to how vendors are charging for agentic AI — whether it’s subscription-based, usage-based, or result-based — and carefully assess long-term financial implications.
Agentic AI represents an exciting new chapter in the AI landscape. By moving beyond reactive systems and embracing autonomous, goal-driven agents, businesses have the potential to unlock unprecedented efficiencies, insights, and agility. But as with all powerful tools, success will depend on thoughtful planning, strong governance, and clear-eyed evaluation of both risks and rewards.
