Written by Lauren Anderson, Dora Peterfi and Aanya Gadde
Artificial Intelligence (AI) is increasingly recognised as a transformative force in procurement, offering significant efficiencies and strategic advantages. However, despite its potential, AI adoption in the sector remains cautious and uneven. While AI's ability to automate routine tasks is well established, its application in more complex, strategic decision-making processes continues to face skepticism.
At Procurato, we view AI as a powerful enabler that can optimise procurement efficiency by automating routine, time-consuming tasks, allowing professionals to focus on more strategic activities that require human expertise. To better understand the current state of AI adoption in procurement, this article draws on university research sponsored by Procurato, which includes in-depth interviews with senior professionals in procurement and management roles within the financial sector. The findings underscore the limited integration and perceived impact of AI, revealing the challenges that hinder broader adoption.
Finally, we highlight specific use cases, such as spend analysis and data enrichment, where AI has already demonstrated practical benefits. We believe that by leveraging AI effectively, organisations can unlock new efficiencies, reduce operational costs, and drive more strategic outcomes in procurement.
Limited adoption of AI technology due to widespread skepticism among Chief Procurement Officers (CPOs)
The adoption of Artificial Intelligence (AI) in procurement is currently marked by limited integration. Despite its recognised potential to transform procurement processes, AI technology is not yet widely embraced. This section delves into the reasons behind the cautious approach of CPOs towards AI, revealing key insights into its current usage, perceived impact, and underlying concerns.
Current usage is limited:
Only around 50% of interviewed professionals reported some level of AI integration in their purchasing processes. The other half either have not adopted AI or are in the very early stages of exploration. This suggests that AI is not fully established and remains in pilot mode for many organisations, signalling the early stages of AI adoption in the procurement sector.
Perceived impact minimal as current role is solely complementary:
Our sample group recognised AI in procurement for its ability to automate routine tasks like data entry and invoice processing but expressed limitations in its ability to effectively handle complex, strategic decision-making. For instance, intricate initiatives such as selecting the best suppliers based on long-term value, managing supplier relationships to mitigate risks, and negotiating contracts that align with broader business objectives require deep market insights and the ability to forecast trends. AI tools like ‘Globality’ and ‘Ariba’, as mentioned by our interviewees, struggle with such tasks since they demand critical thinking and deep understanding of market dynamics. The limited integration into executive decision-making processes signifies the technology’s current stage and emphasises the ongoing need for human intervention. Therefore, AI serves as a complement to human roles, supporting but not replacing critical functions.
CPOs are skeptical about AI’s reliability:
Although there is recognition of AI's potential, the overall sentiment is that AI is not yet reliable or mature enough to be fully trusted and it lacks the sophistication needed to drive significant change in the purchasing process. One interviewee expressed concerns, noting that AI might produce "irrelevant or misleading outputs" in certain contexts. Others are concerned about the ethical issues in decision-making associated with AI and fear that its adoption may be more about keeping up with competitors rather than its genuine utility. Some CPOs are also concerned that senior management in other departments may not fully understand the technology they are acquiring, which poses risks as they might not be aware of the true capabilities and limitations of the AI solutions they are investing in.
Barriers and challenges
So far, our research has shown that the limited adoption of AI in procurement is often rooted in a fundamental lack of understanding among senior management. Criticisms that they "do not have a clue of what they are buying and their liabilities" highlight a significant knowledge gap that contributes to this hesitation. Without a clear grasp of AI's benefits and risks, decision-makers tend to approach the technology with caution. Furthermore, our interviews revealed that some CPOs are driven by a "Fear of Missing Out" (FOMO) rather than a deep understanding of AI's potential. This reactive stance can lead to superficial adoption, where AI is implemented more as a competitive catch-up than a strategic tool to meet specific organisational needs.
In addition to these concerns, our research has identified three critical barriers that CPOs view as significant obstacles to implementing AI in their procurement processes.
Stakeholder and supplier engagement:
The purchasing process in the financial sector is characterised by complex stakeholder and supplier engagement, where finding and shortlisting suitable suppliers is challenging due to niche requirements and lengthy procedures. This complexity demands significant manual effort and reliance on experienced professionals, which can deter AI adoption. The nuanced and dynamic nature of these interactions requires a level of contextual understanding and judgment that current AI tools may struggle to replicate [McKendrick and Thurai (2022), (Shine, 2023)]
Complexity of financial budgeting and due diligence:
Budgeting in large, complex financial environments, along with conducting due diligence, poses significant challenges. Financial budgeting often involves detailed calculations and considerations, while due diligence requires thorough analysis of data and financial conditions. The sophisticated and multifaceted nature of financial data necessitates human-driven analysis to navigate lengthy budgeting scenarios and due diligence processes, limiting the perceived utility of current AI solutions, which are often seen as insufficiently advanced for these tasks. As one of our interviewees put it: “Purchasing remains a largely manual process, with elements of technology employed (on-line forms, etc.,) to address basic functionality issues. Beyond this, the process remains staff intensive and reliant on skilled and experienced purchasing professionals. “
Regulatory and ethical concerns:
CPOs are particularly concerned with compliance with strict regulations and ethical implications of AI adoption across the business, especially regarding data protection and decision-making biases. One of the interviewees expressed these concerns, stating: “There are concerns about AI taking control of the business, as it may act in ways you don’t want. Senior leaders often lack understanding of what they’re purchasing and the associated liabilities. Furthermore, without the right corporate experts in the room asking the right questions, it’s a recipe for disaster.” These concerns add complexity and slow technology adoption, as organisations hesitate to fully integrate AI until it aligns with regulatory standards and ethical norms. Based on our research, among the main concerns are data privacy, accountability and transparency. For instance, data protection regulations like GDPR emphasise fairness and transparency, yet AI systems often face challenges in ensuring unbiased decision-making, which can lead to discrimination and legal complications (European Parliament, 2020). Additionally, as AI algorithms process vast amounts of data, the potential for embedded biases becomes a significant issue, potentially leading to unfair outcomes in areas like recruitment or finance (CompTIA, 2022). These issues signify the need for rigorous oversight and the development of clear guidelines to ensure AI is used responsibly within procurement processes.
Procurato’s perspective: strategies for organizations and practical use cases
At Procurato, we see AI as a game-changing tool that, while still in the early stages of adoption within procurement, holds immense potential to revolutionise not only the financial sector but industries across the board. We see AI as a powerful enabler that can optimise efficiency by automating routine and time-consuming tasks, such as administrative duties. This allows procurement professionals to shift their focus towards more tactical and impactful activities that require human expertise, such as market analysis and long-term supply chain planning.
Admittedly, there are constraints and obstacles in the current landscape of AI implementation in business activities. However, we believe that these challenges can be addressed through thoughtful and deliberate integration of AI technologies. A balanced approach is essential, where AI enhances human capabilities and decision-making processes without replacing the critical judgment and deep understanding that experienced professionals bring to the table. In the following section, we will explore specific use cases where we have observed practical benefits of AI in procurement.
Spend analysis
AI-driven models have revolutionised spend analysis by automating the categorisation and analysis of spending data. Machine Learning (ML) and AI technologies can efficiently process and categorise large volumes of spend data, significantly reducing the time and effort previously required for manual analysis. However, while this automation accelerates the spend analysis process and enhances accuracy, it is not without its challenges. Successful implementation requires access to massive datasets that have been accurately trained, as well as a robust audit process to ensure reliability. As such, AI serves as a powerful enabler in spend analysis but is not yet a complete solution on its own. Organisations must still invest in the right resources and oversight to fully harness its potential for deeper insights and more informed procurement strategies.
Data enrichment
AI also excels in data enrichment by extracting valuable information from large and complex files that are often stored outside traditional procurement systems. For example, technologies like Textmine automate the extraction and structuring of critical terms from extensive document databases, converting previously inaccessible information into a more organised and searchable format. This capability is crucial for enhancing risk management and refining procurement data insights. By digitalising and integrating such information, AI facilitates more effective category management and process automation. Enriched data offers a more comprehensive view of supplier information and market conditions, ultimately supporting better strategic decision-making and improving operational efficiency. As AI technologies continue to advance, they are set to play an increasingly vital role in streamlining procurement processes and delivering deeper, actionable insights.
Writing RFPs and analysing contracts
Another emerging application of AI in procurement is in the drafting of Request for Proposals (RFPs) and the analysis of contracts. AI tools are increasingly being utilised to generate and customise RFP documents, streamlining what has traditionally been a time-intensive process. Not only are these AI systems helping procurement teams craft precise RFPs, but suppliers are also beginning to interact with AI-driven platforms, enabling automated responses and communications. This automation improves the efficiency and quality of the RFP process, ensuring that procurement decisions are well-informed and aligned with business objectives.
Moreover, AI is proving to be a valuable asset in contract analysis. Several companies have reported success in using AI to identify critical clauses, gaps, and inconsistencies within contracts. By leveraging Natural Language Processing (NLP) algorithms, AI systems can quickly scan and interpret vast amounts of contractual language, flagging potential risks and ensuring compliance with organisational standards and legal requirements. This capability significantly enhances the contract management process, reducing the likelihood of costly oversights and enabling more strategic contract negotiations.
Conclusion
In conclusion, AI is set to play an increasingly critical role in procurement, offering numerous benefits from spend analysis and data enrichment to the automation of RFP creation and contract analysis. While AI is not yet a total solution and requires significant oversight and resources, its potential to drive efficiencies, enhance decision-making, and support more strategic procurement processes is undeniable. At Procurato, we believe that by embracing these technologies thoughtfully, organisations can unlock new levels of performance and value in their procurement functions.
References
Finn, T. and Downie, A. (2024). AI in Procurement | IBM. [online] www.ibm.com. Available at: https://www.ibm.com/think/topics/ai-in-procurement.
McCrea, B. (2024). Procurement Approaches Gen AI with a Skeptical Eye. [online] Supplychainconnect.com. Available at: https://www.supplychainconnect.com/procurement/article/55017862/procurement-approaches-gen-ai-with-a-skeptical-eye [Accessed 19 Aug. 2024].
Perry, N. (2023). Why 2024 Will Be Pivotal for AI in Procurement. [online] Available at: https://procurementmag.com/articles/why-2024-will-be-pivotal-for-ai-in-procurement.
www.mckinsey.com. (n.d.). Gen AI value in procurement. [online] Available at: https://www.mckinsey.com/capabilities/operations/our-insights/operations-blog/making-the-leap-with-generative-ai-in-procurement.
McKendrick, J. and Thurai, A. (2022). AI Isn’t Ready to Make Unsupervised Decisions. [online] Harvard Business Review. Available at: https://hbr.org/2022/09/ai-isnt-ready-to-make-unsupervised-decisions.
Shine, I. (2023). These are the jobs that AI can’t replace. [online] World Economic Forum. Available at: https://www.weforum.org/agenda/2023/05/jobs-ai-cant-replace/.
European Parliament (2020) The Impact of the General Data Protection Regulation (GDPR) on Artificial Intelligence. Available at: https://www.europarl.europa.eu/RegData/etudes/STUD/2020/634452/EPRS_STU(2020)634452_EN.pdf
CompTIA (2022) Common Ethical Issues in Artificial Intelligence. Available at: https://connect.comptia.org/blog/common-ethical-issues-in-artificial-intelligence
Software AG (2023) The Data Integration Advantage: Scalable AI. Available at: https://www.softwareag.com/en_corporate/resources/data-integration/wp/data-integration-scalable-ai.html
Comments