Let’s get into this week’s brief.

Sorry for missing last week. I spent a week in deep work on AI Agents and more.

Data Insights: The brief draws on 120 distinct sources, spanning market research reports, analyst forecasts, government policy documents, industry surveys, academic studies, press releases, journal articles, white papers, and technical preprints.

Global generative AI spending will surge to $644 billion in 2025, representing a 76% year-over-year increase, according to new Gartner data. This massive investment surge is primarily driven by vendor infrastructure buildout rather than enterprise custom development, with companies shifting from building in-house AI tools to purchasing off-the-shelf solutions.

The trend reflects a fundamental change in procurement strategy: "CIOs are no longer building generative AI tools, they're being sold technology", creating new opportunities for procurement teams to negotiate better vendor packages as supply increases.

The Hackett Group's 2025 study reveals that procurement workloads are expected to increase by 9.8% in 2025, while headcount and budgets are projected to see only marginal growth, resulting in an unprecedented 8.8% efficiency shortfall. This productivity crisis is forcing procurement leaders to accelerate automation investments, with 89% relying on automation to reduce manual tasks3 and 86% of procurement teams reporting an increased workload despite stagnant headcount. 3 The gap between rising expectations and static resources is becoming the defining challenge of 2025.

New survey data shows 24.6% of procurement professionals use ChatGPT daily, with 198 out of 804 surveyed professionals claiming to use it every working day. The tool has generated over 3,000 procurement-related emails and is being utilised for a wide range of purposes, from contract analysis to supplier research. Notably, the adoption of ChatGPT boosts regulatory compliance by 18-20% and reduces dispute-induced delivery delays by 4-5 days, resulting in annual savings of over $100,000 for mid-sized enterprises. However, 25% of procurement professionals still admit to never having used ChatGPT, indicating a massive untapped potential.

Prime Minister Keir Starmer announced plans to save £45 billion in public service spending through AI and digital transformation, with the government targeting 10% of civil servants working in "tech and digital roles" within five years. The initiative includes 2,000 additional tech apprenticeships by 2030 and deployment of AI tools across government departments. Early tests show an AI call centre helper halved response times for complex queries, demonstrating the potential for similar productivity gains across procurement functions.

Industry analysis reveals that 60-70% of daily tasks performed by lower-level procurement professionals will be automated in the near future, with automation already enabling 40% improvements in transaction speeds and significant operational cost reduction. Companies implementing AI-driven procurement report productivity gains of 30-50% across business processes, while AI applications in risk assessment and credit scoring have reduced processing times by approximately 40% and improved approval rates by 25%. The data suggests we're approaching an inflection point where automation becomes essential for competitive procurement operations.

These developments signal that 2025 is the year procurement automation moves from optional to essential. The combination of massive vendor investment, proven productivity gains, and government leadership is creating an environment where AI adoption becomes a competitive necessity rather than an experimental advantage.

The message is clear: procurement teams that don't embrace AI automation in 2025 risk falling into an efficiency gap that will be increasingly difficult to bridge as competitors gain sustainable productivity advantages.

*The brief draws on 120 distinct sources spanning market-research reports, analyst forecasts, government policy documents, industry surveys, academic studies, press releases, journal articles, white papers, and technical preprints.