Artificial Intelligence: The Impact on Public Procurement - Opportunity & Risk

By now, most are beginning to view artificial intelligence as something to be reckoned with. Regarding public procurement, there are two main issues: What should one know about purchasing AI-type systems, and how can AI be used to streamline and improve the overall procurement function?

As with any new technology, the first issue can be addressed by knowing about the vendor, how security and privacy can be guaranteed, and how to negotiate a reasonable price for the services provided. Nearly every vendor claims its products and services are “AI-powered,” which is often overblown and misleading. One should never be afraid to ask, “What do you mean by “AI-powered” and what makes this product or service different from previous offerings? In many cases, “AI-powered” merely explains or promotes new features that rely on improved search abilities and anticipating user’s needs. Remember, not all AI is the same. For example, most people have come to know AI through ChatGPT, Microsoft Copilot, or Google Gemini. These relatively new entrants utilize generative AI.  Generative AI refers to artificial intelligence systems designed to create new content, such as text, images, music, or video, by learning from existing data. Common examples include AI tools for generating realistic images or writing human-like text.  The most exciting opportunities lie with how AI, in general, can assist in the procurement process.

Here are just a few areas in which AI can assist public procurement:

1. Automated Supplier Evaluation: AI can analyze large datasets to evaluate suppliers based on cost, quality, delivery performance, and compliance with regulations. This helps in selecting the most suitable suppliers.
2. Predictive Analytics: AI can forecast demand and trends, helping procurement managers make informed decisions about future purchases and inventory management. Predictive models can anticipate shortages, price fluctuations, and optimal order times.
3. Contract Management: AI can automate contract creation, track compliance, and flag deviations. Natural Language Processing (NLP) can extract key contract terms and clauses and ensure they meet regulatory standards and organizational policies.
4. Spend Analysis: AI systems can analyze spending patterns to identify cost-saving opportunities, detect anomalies, and ensure compliance with budgetary constraints. This helps with better financial planning and management.
5. Fraud Detection: Machine learning algorithms can identify suspicious activities and patterns that may indicate fraud, such as unusual bidding behavior, conflicts of interest, or invoice discrepancies.
6. Supplier Relationship Management: AI can monitor supplier performance and provide insights into areas for improvement. It can also predict potential risks in the supply chain, such as supplier bankruptcy or geopolitical issues, and suggest mitigation strategies.
7. Process Automation: AI can automate routine tasks such as invoice processing, purchase order creation, and approvals - freeing procurement professionals to focus on strategic activities.
8. Enhanced Transparency and Accountability: AI can ensure that procurement processes are transparent by maintaining detailed logs and records of all transactions and decisions. This can help with audits and maintaining public trust.
9. Smart Bidding Platforms: AI-powered platforms can facilitate audits and bidding processes by matching procurement needs with the most suitable bidders and optimizing bid evaluations based on multiple criteria.
10. Sustainability and Ethical Sourcing: AI can track and ensure procurement practices adhere to sustainability and ethical standards by monitoring supplier practices, certifications, and environmental regulations.

Procurement professionals have begun experimenting with AI applications, and it is too early to evaluate how well these new tools are performing - though early adapters are expressing both confidence and optimism. However, as with any new technology, one must always be prepared for risks. What follows is a list of areas of concern:

1. Data Quality and Availability: AI systems require large amounts of high-quality data to function effectively. In many cases, procurement data is fragmented, inconsistent, or incomplete, which can hinder the performance of AI algorithms.
2. Integration with Existing Systems: Integrating AI solutions with existing procurement and enterprise resource planning (ERP) systems can be complex and costly. Legacy systems may not be compatible with new AI technologies, requiring significant upgrades or replacements.
3. Regulatory and Compliance Issues: Strict regulations and compliance requirements often govern procurement processes. Ensuring that AI systems adhere to these regulations and can adapt to changes in compliance standards is a significant challenge.
4. Transparency and Explainability: AI algorithms, especially complex ones like deep learning, can be opaque and difficult to understand. This lack of transparency can be a barrier to adoption, as procurement professionals may be reluctant to trust decisions made by "black box" systems.
5. Bias and Fairness: AI systems can inadvertently perpetuate or even exacerbate biases present in the training data. Ensuring that AI-driven procurement processes are fair and unbiased is crucial, especially in public procurement, where equity and fairness are paramount.
6. Cybersecurity Risks: AI systems can be vulnerable to cyberattacks, which can compromise sensitive procurement data and disrupt operations. Ensuring robust cybersecurity measures are in place is essential to protecting AI systems and data.
7. Change Management and Resistance to Adoption: Implementing AI in procurement requires changes in processes, workflows, and potentially job roles. Resistance from employees due to fear of job displacement or lack of understanding of AI benefits can hinder adoption.
8. Cost and Resource Constraints: Developing and deploying AI solutions can be expensive, requiring significant investments in technology, infrastructure, and skilled personnel. Budget constraints can limit organizations' ability to implement AI in procurement.
9. Ethical Concerns: The use of AI raises ethical questions related to privacy, surveillance, and the potential misuse of data. Addressing these concerns is essential to maintain public trust and ensure the ethical use of AI in procurement.
10. Continuous Improvement and Maintenance: To remain effective, AI systems require continuous monitoring, updating, and maintenance. This ongoing effort can be resource-intensive and requires specialized skills to ensure the system evolves with changing needs and environments.

We all have much to learn and share, and I hope this short piece helps public procurement professionals better navigate what lies ahead.