Our CEO, Ian Makgill, recently discussed imminent AI impacts on public procurement in an enlightening The Trowers podcast. Rapid advances in generative language models like ChatGPT foreshadow revolutionary change ahead. Yet governments have scarcely begun responding, given the swift pace of progress. Ian explored both challenges and opportunities from AI alongside transparency and data improvements still needed in the ecosystem.
Overhaul Needed in Text-Heavy Tender Practices
Lengthy written bid requirements impose unnecessary burdens, likely exacerbating issues as AI assistance becomes pervasive. Rather than resisting this, buyers should streamline processes:
- Curtail verbose specifications.
- Simplify evaluation while maintaining strict standards.
- Help level the playing field.
Short, targeted questions will only be accessible for unaided software.
Risks of AI Evaluation Overstated
AI cannot drive the consensus to agree on requirements or determine successful awards. Human discretion and explanation remain essential in public procurement. Some analytical use cases, like correlation studies on tender criteria variations with outcomes, hold promise. However, limited scope exists for AI to directly conduct or decide evaluations currently and foreseeably.
Transparency & Data Improvements Must Precede Prediction
Ian highlighted the immense potential of analytics once enough high-quality open data accumulates. Linking detailed tender specifications to eventual contract success could underscore best practices. However, sound analysis requires information gaps and deficiencies to be resolved through publication diligence. Until then, predictive procurement optimisation remains speculative rather than achievable.
Procurement Leaders Must Address the Knowledge Gap
Given the rapid pace of AI change, procurement professionals urgently need to educate themselves on state-of-art capabilities, prudent applications and ethical risks requiring vigilance. Many industries now actively harness predictive modelling and automation fueled by collected data. Public procurement lags – but cannot hide from technological forces reshaping society. Proactive assessment and planning are overdue.
With more data standardisation, open availability and subsequent maturation of analytical techniques, AI and machine learning will impart radical shifts across public procurement. Exact manifestations remain difficult to anticipate this early, but underprepared organisations likely struggle with disruptive impacts. We must continue advancing data foundations while staying cognizant of AI progress on the horizon.
The focus is on AI’s transformative potential alongside transparency and policy prerequisites needed to apply these technologies in the future prudently.
Spend Network holds government procurement data for more than 150 countries and 700 sources globally. We work hard to collect, organise and analyse procurement data at scale for export agencies and businesses. Subscribe to our newsletter and stay informed about the latest news in the industry and how our data can help you.