Procurement AI is coming to getcha

Procurement teams across the globe have embraced AI with all the enthusiasm of a teenager wielding a chainsaw - excited, slightly reckless, entirely sure they’ll “figure it out”, though they might lose a limb or two.    

Gone are the days when requisition forms were lovingly hand-mailed, or, for more advanced organisations, faxed with abandon to nowhere in particular. Now we have predictive analytics that can guess what we wanted to buy, just after we’ve already bought it.  

AI promises to revolutionise procurement, replacing human intuition with elegant algorithms that have never met a supplier, experienced panic before a quarterly review, or been ghosted by an RFQ. Negotiations are now handled by neural networks trained on the finest customer service transcripts, including such classics as “Can I speak to your manager?” and “No, I will not turn it off and back on again.”  

But don’t worry: procurement professionals aren’t obsolete. They’re just “strategically repositioned” to monitor the bots who have taken their jobs. It’s collaborative. Really. Like a relay race - except the baton is a 400-page contract and the finish line is perpetually moved by agentic AI.  

AI in Procurement: A Brief History of Hype and Selective Ignorance  

From invoice automation to algorithms that can predict that your supplier will let you down at precisely the worst moment, AI's rise in procurement is less revolution and more politely procrastinated evolution.   

Yes, we’ve all heard that in a mere decade the procurement professional will be an endangered species. Unfortunately, extinction has been delayed by budget approvals and ERP integration woes.  

Automation, Intelligence, and Other Corporate Buzzwords  

Back in 2018, AI was just the intern who sorted invoices faster than Rachel from Accounts (at the Bank of England). By 2025, it had graduated to a semi-useful assistant-good at spotting spend anomalies but still utterly baffled by the nuance of “buy cheap, buy twice.”   

Now it’s being tasked with strategic sourcing, negotiation modelling, and interpreting contracts longer than a Tolstoy novel. God help us. Rachel won’t.  

Investment Priorities: Procurement’s Place in the Corporate Pecking Order  

Let’s face it: if corporate investment were a family dinner, procurement would be at the kids’ table with a huge bowl of Wotsits.  Sales, marketing, even facilities management - all get their tech treats first. Procurement?   We get a screen cleaner for the ZX Spectrum.  

Generative, Predictive, Agentic… Oh My!  

The lexicon now reads like something cooked up in a sci-fi writer’s sauna. Agentic AI drafts strategy, generative AI executes it, and Predictive AI tries to guess what went wrong-none of them have realised that the crazy people in logistics order parts based on lunar cycles.  And yes, 98% of procurement people don’t know the difference between generative and agentic.  

From Spend Analysis to Strategic Annihilation  

NLP and ML are now doing the heavy lifting in spend categorisation, risk assessments, and compliance checks. The bots (or is that Diella from Albania) are reading contracts, scoring suppliers, and generally making procurement professionals look suspiciously like spare parts. Strategic roles? Soon absorbed into the amorphous “commercial function,” the HR euphemism for “everyone’s been automated except for, er, HR.”  

The Two-Speed Evolution: Fast Lane for Pharma, Slow Lane for Public Sector 

Let’s be honest: industries that actually care about procurement are zooming ahead. For the rest, AI’s promise is still trapped in a PowerPoint from 2019. We’re witnessing a two-speed adoption curve: one lane heading toward the future, the other stuck behind a procurement director who thinks “AI” stands for “Artificial Insemination” or possibly “Another Initiative.”  

Final Thoughts: Are We Uploading or Uplifting?  

In ten years, we may well be in the Matrix - our procurement souls forever digitised, our thoughts validated only if machine-learned. Until then, procurement bloggers everywhere will cling to relevance with LinkedIn think-pieces, while actual bots quietly take over their audience one algorithmic hot take at a time.  

So, put the chainsaw down, let HAL 9000, or Skynet or maybe Wall-e take over.