In recent years, the procurement sector has undergone a formidable transformation. The procurement function has gradually become digitalised, automated and has gained a strategic position within businesses. With the development of Artificial Intelligence, this dynamic continues. After predictive AI and generative AI, we are entering a new era: that of agentive AI. This technology consists of acting proactively, within a defined framework. For procurement, this promises ever more efficiency, agility and strategic impact.
What is agentive AI?
Agentive AI is designed to act autonomously, without human intervention. This technology can identify objectives, break them down into different tasks and execute processes. All this with minimal human oversight.
These are what we call AI agents. These software programmes are capable of planning, adapting and interacting with their environment. Moreover, they can learn from data and improve their performance over time.
Until now, procurement has relied on robotic process automation (RPA) to execute repetitive tasks, on predictive AI to anticipate risks and on generative AI to produce content. Now, agentive AI capabilities go even further through its ability to act and reason autonomously. According to Gartner, 33% of enterprise software applications will include agentive AI models by 2028, enabling 15% of daily work decisions to be made autonomously.
Thanks to this new type of agent, procurement teams can thus automate complex use cases comprising several stages and achieve better results. This is all the more interesting, as there are still many repetitive tasks and data manipulation in procurement roles today. Santosh Nair, chief product officer at GEP, emphasises: "AI agents are set to revolutionise procurement by addressing existing challenges, enhancing efficiency, and paving the way for a more strategic and proactive procurement function. Organisations that embrace this technology stand to gain a competitive edge in the evolving business landscape."
Its strategic applications in procurement
AI agents can intervene throughout the procurement value chain, from expenditure management to invoicing. Through three key examples, here is an overview of their formidable potential on the Source-to-Contract process.
Sourcing and supplier evaluation
Agentive AI can generate and evaluate requests for proposals completely autonomously. When autonomous agents detect a deviation from contractual terms, they can automatically trigger a request for proposal (RFx). They adapt the content, select the suppliers to approach and evaluate their responses using a dynamic scoring system. During this process, they can adjust elements based on external factors. For example, if market conditions change or new regulatory constraints arise.
Risk management
Until now, risk management has always relied on periodic assessments or alert systems beyond certain predefined thresholds. Tomorrow, AI agents can also serve as alert systems by continuously monitoring large flows of structured or unstructured data (supplier performance, financial reports, news feeds...). They are able to detect weak signals and flag anomalies but also suggest solutions. This allows teams to anticipate problems and guarantee business continuity and their company’s resilience, despite the unforeseen.
Contract negotiations and contractualisation
Agentive AI provides valuable support for contract negotiations. It will be able to carry out scenario analyses, simulate negotiation results based on data, and even draft bespoke contractual clauses. Agentive AI assists procurement teams conducting negotiations by providing them with suggestions and risk assessments in real time. In some cases, AI agents will even negotiate completely autonomously. This is explained by Christoph Menne, partner at German management consultancy firm absolut Group: "We’re seeing customers in Europe who are starting to leave the sourcing and procurement process for low-value items and services completely to agentic AI"
The challenges for implementing agentive AI
Whilst agentive AI offers formidable potential, it also raises its share of challenges from a technical, cultural and organisational standpoint.
Infrastructure and data
Infrastructure and data integration are crucial for implementing agentive AI. Companies must rely on modern and interoperable infrastructure to connect the entire procurement ecosystem (ERP, SRM, etc.). Beyond technology, data quality and governance are also key. Fragmented, obsolete or unreliable data will invariably impact AI agents’ ability to detect signals, make decisions and act autonomously and securely.
Manutan’s advice
E-procurement plays a key role in this transformation. By centralising procurement flows within a single platform, companies structure data, improve its quality and create the necessary conditions for advanced automation. An essential prerequisite for fully benefiting from AI agents.
Manutan’s “Integration / eProc / Savinside” service handles integrations (ERP, e-procurement), with rapid connection, training and dedicated support to optimise procurement and improve efficiency. (available in Belgium, the Czech Republic, Denmark, Sweden, Finland, France, Germany, Hungary, Italy, the Netherlands, Norway, Poland, Slovakia, Spain, Switzerland, the United Kingdom, Portugal, at date of content publication)
Change management
Agentive AI requires a major cultural transformation. Teams will have to accept entrusting certain tasks to Artificial Intelligence, but also understand and master these autonomous agents. Such a scenario can raise fears, resistance and questions. To prepare the ground, it is advisable to rely on clear communication, training programmes and management support. This is what Ian Tickle, Chief of Global Field Operations at Freshworks, advocates: "AI does not replace procurement teams. It frees up time, reduces low value-added tasks, and allows professionals to refocus on strategy, negotiation, and supplier relationships. It is still necessary to support this change: train teams, create test environments, and build a culture of experimentation."
Control and compliance
Lastly, one of the major challenges lies in the level of autonomy granted to AI agents. When an AI can act autonomously, the question of decision-making responsibility becomes central. Companies must therefore define a clear governance framework that specifies human supervision mechanisms, decision explicability as well as compliance with regulatory constraints and internal policies. This is how companies will be able to guarantee the traceability of actions but also alignment with business, ethical and regulatory priorities.
Agentive AI systems mark a new key stage in the transformation of procurement. Without further delay, procurement departments have every interest in experimenting with it, through a pragmatic approach. To do this, they will need to start by selecting the right use cases and technology partners. This is a first step towards a more agile, efficient procurement function that is better prepared for the future.
