2. The 5 Procurement Problems AI Notices Immediately

Series: What AI Thinks of Procurement — Post 2

If Post 1 proved anything, it’s that AI sees procurement very differently from humans.
Where humans see “a busy Tuesday,” AI sees patterns, red flags, risk signals, and — occasionally — chaos wearing a polite business-casual outfit.

So today, we let AI sit in the driver’s seat and share the five things it notices the instant it looks at most procurement functions.

You may want to sit down. Or hydrate. Or emotionally prepare.

Let’s go.

1. Siloed Data — AKA: “Why does everyone have their own spreadsheet?”

AI takes one look at procurement data and immediately says:

“Why is the supplier master in 12 places? Why does Finance have a different version? Why is Category Management using a Google Sheet last edited during the Bronze Age?”

Humans call this “siloed data.”
AI calls it “a scavenger hunt with no treasure.”

Here’s what AI doesn’t understand (but humans know too well):

Systems don’t talk to each other
Teams don’t talk to each other
And honestly… no one wants to open that old shared drive folder named “PROCUREMENT_FINAL_v27.xlsx”

AI’s reaction:
“This could all be avoided with integrated data, APIs, or literally anything from this century.”

2. Manual Approvals — AKA: “The approval queue nobody escapes from.”

AI is extremely confused by manual approval chains.

It sees:

8 people approving a $300 purchase
Approvers on vacation until 2049
A VP approving office chairs even though they have never sat in one

From AI’s perspective, manual approvals create:

Delays
Bottlenecks
Audit nightmares
And a lot of “Hey, circling back!” emails

AI’s solution?
Automate 70% of approvals and reserve human decision-making for the stuff that really matters, like “Should we sign a three-year SaaS contract?” or “Why did this supplier just double their price without telling us?”

AI’s thought bubble:
“Why are humans approving things they don’t even want to approve?”

3. Fragmented Supplier Information — AKA: “Same supplier, 4 names, 3 IDs, 2 addresses, and 0 consistency.”

If you ever want to see an AI have a meltdown, show it a supplier master file.

It finds:

Duplicate suppliers
Different spellings of the same supplier
Parent companies that no one realized were connected
Terms hidden in contracts no one remembers signing

And the best one:

One supplier with three different emails, one of which is definitely from 2006.

AI’s suggestion?

“Unify supplier data. Clean it. Normalize it. Then maybe — just maybe — the universe will make sense.”

To AI, fragmented supplier info creates:

Bad reporting
Inefficient sourcing
Unreliable risk assessments
Endless confusion

To humans, it creates headaches.
To AI, it creates existential disappointment.

4. Reactive Risk Management — AKA: “Everyone waits until something is on fire.”

AI sees risk forming months before it happens.
Humans, however, often find out about risk when:

A supplier misses a delivery
A shipment disappears into the void
A regulatory issue pops up
Someone says, “Did anyone know this supplier is being sued?”

AI monitors:

News sentiment
Performance trends
Financial signals
Delivery patterns
Weather disruptions
Geopolitical noise
Industry volatility

Humans monitor:

“Wait… why is this late?”

AI’s critique is that procurement often reacts after risks hit instead of preventing them.

AI’s simple philosophy:
“If you can predict it, you can prevent it.”

Radical, I know.

5. Underutilized Analytics — AKA: “The dashboards are pretty, but no one is using them.”

AI loves analytics.
It’s basically analytics’ biggest fan.

So when AI sees:

Beautiful dashboards
Custom reports
Trend charts
Category data
Supplier scorecards

…and learns that nobody checks them?

AI is devastated.

Analytics today are often used like:

A safety net
A quarterly report prop
A “just in case leadership asks” file

But AI wants analytics to be:

Predictive
Prescriptive
Embedded in decisions
Driving recommendations
Informing strategies

AI’s feedback:
“Use the data. Trust the data. Stop printing the dashboards — I made those interactive on purpose.”

Final Thoughts: AI’s Procurement Pep Talk

To summarize what AI sees instantly:

Siloed data → “Why is everything everywhere?”
Manual approvals → “Let the robots help, please.”
Fragmented suppliers → “One supplier, many identities.”
Reactive risk → “Predict before panic.”
Underutilized analytics → “You have the answers — use them.”

AI isn’t judging.
Okay, maybe a little.

But mostly, it sees how close procurement is to transforming into a high-performing, insight-driven, strategically fearless machine.

And if AI could leave procurement with one final note, it’d be this:

“The problems aren’t permanent — but the opportunities are endless.”