Skip to content
Exclusive partner BME Symposium18–19 November · Berlin · Booth A16Book a meeting
Contact

Autonomy is a dial, set by reversibility.

The question is not whether AI may decide in procurement. It is which decision, on what evidence, and who notices when it gets it wrong.

In short

AI governance in procurement defines which tasks AI may take on, which data it works with and where people approve. With agentic AI this belongs in the architecture: autonomy per process step, set by reversibility and materiality, approval points, data sovereignty and traceable decisions. Without an approved path, employees use their own tools, and control is lost.

Data: third-party research with source. Our view: the joint position of H&Z and Elevatry, indicative observations from project work. Model: calculations and examples, not customer results.

Autonomy is not one policy for everything. It is a dial per process step.

Two criteria do most of the work: how easily an error can be undone, and how much is at stake. A misclassified order is corrected with one click. An award binds the company for years. Both errors cost something, but not the same thing.

That gives a dial with three levels. In autonomous mode a Companion completes an entire process where errors are cheap and reversible; people approve the result or check samples. In guided mode the Companion prepares and the team confirms every step; that applies to evaluations, scenarios and anything material and hard to undo. In orchestrated mode, currently in development, several Companions work together. Awards and terminations are decided by a person at every level.

Gartner expects over 40% of agentic AI projects to be cancelled by the end of 2027. A common reason is that nobody defined what the machine may decide. The pilot then stays a pilot, or scales without control. Both are expensive.

Figure 1Our viewThe harder a decision is to undo, the less autonomy it gets.
Autonomous, sampled
Autonomous, sampled
Autonomous, with approval
Guided
Human decides
Spend classificationDeadline monitoringTender preparationBid evaluationSupplier award

View of Elevatry and H&Z. Typical starting levels for a first deployment; each organisation calibrates its own thresholds.

Autonomy is earned with evidence, not granted by decree.

Measure Companions the way you measure suppliers. For each process step, track acceptance, override and error rates. A step that runs a quarter in guided mode with few overrides switches to autonomous mode. If overrides rise, it goes back to guided mode.

The second risk is quieter: polished output approved without challenge under time pressure. The design response is evidence behind every claim, a named person responsible for challenging, and sampled reviews even at higher levels. Whoever only approves what they could no longer produce themselves is no longer reviewing. They are signing.

Elevatry records every recommendation, approval and correction with source, person and time in the Event Trail. That is the basis for justifying autonomy, and it allows EU AI Act classification per use case rather than one verdict for the whole platform.

>40%of agentic AI projects are expected to be cancelled by the end of 2027.DataGartner, June 2025
A step that ran a quarter with almost no overrides has earned autonomous mode.

Rent the model. Own the mandate.

According to Microsoft, 71% of AI users in German companies use their own, non-approved tools. Contracts and supplier prices then end up in private accounts with no data processing agreement. Bans only push that use into channels nobody sees. What helps is an approved path that works better day to day.

Sourcing that path calls for procurement logic. No procurement team would single-source a critical component without a contingency. Many AI strategies do exactly that with their intelligence. Models are converging and available to every competitor. Equating strategy with model choice means competing on what everyone can buy.

What you should own is method, memory and mandate. The Model Gateway selects the model per task. The Company Brain stays separate from the model in the EU data zone, with no training of shared models on your data. If the provider changes, the context stays.

71%of AI users in German companies use their own, non-approved AI tools.DataMicrosoft Work Trend Index 2024

Companions for this topic

What Elevatry does for you.

Companions work around the clock in your data, with the H&Z method from more than 5,000 procurement projects. They prepare decisions, work in your experts' feedback and, once you approve, carry out the next steps. Your team decides.

All Companions
Model

Index clauses: which level for which step?

A company has 240 contracts with index clauses covering €40M of spend. Today they are checked by hand twice a year. Index reductions are claimed on average five months after they arise, if at all.

  1. 01
    Break the process down

    Matching index movements against contracts is cheap and reversible. A letter to a supplier leaves the building and cannot be recalled. An escalation or termination is material.

  2. 02
    Start cautiously

    In the first quarter every step runs in guided mode. Buyers review each calculation and each letter.

  3. 03
    Collect the evidence

    Of 410 calculations, 6 are overridden, 1.5%. All trace back to contract amendments missing from the system. The data source is fixed.

  4. 04
    Turn the dial

    Calculation switches to autonomous mode with a 10% sample. Letters run in autonomous mode with approval. Escalations are still decided by people.

  5. 05
    The person decides

    The head of procurement sets and owns the levels and thresholds. Every change is justified in the Event Trail.

Index reductions are claimed within one month instead of five. With a 2% index fall on €40M, that is about €0.3M a year in the model. No letter leaves the building without approval.

Illustrative, not a customer result. Assumptions: €40M spend under index clauses, 2% index fall, four months gained (2% × €40M × 4/12 ≈ €0.27M).

Questions worth asking.

  1. 01

    Which procurement decisions may a machine take today, and where is that written down?

  2. 02

    How much of our AI use runs through private accounts, and how do we know?

  3. 03

    Could we switch our AI model tomorrow without losing context and decision history?

  4. 04

    Who in our organisation notices when a poor AI recommendation is approved anyway?

What procurement leaders ask us.

Is AI in procurement compatible with GDPR and the EU AI Act?
Yes, and the assessment is made per use case, not for a platform as a whole. Three things carry the review: data in an EU data zone without training shared models, human oversight at defined approval points, and a complete record of every recommendation, approval and correction. Elevatry produces that record in operation, not after the fact. Documenting this way lets you extend autonomy to more spend. The legal assessment stays with your organisation.
Who is liable if an AI recommendation in procurement is wrong?
The company, and within it the person who approves. AI carries no accountability, so it only prepares hard-to-reverse decisions such as an award. The real risk is a polished recommendation approved under time pressure without challenge. Approving what you can no longer check is signing, not reviewing. The remedy is evidence behind every statement, a name and timestamp on every approval, and spot checks even at high autonomy levels. Your legal team still reviews the individual case.
Can buyers put contracts and supplier prices into public AI tools?
Not without a data processing agreement, yet it happens every day. According to the Microsoft Work Trend Index 2024, 71% of AI users in German companies use their own, non-approved tools. This is not a discipline problem but a supply problem: the buyer needs to understand the index clause by Friday. A ban only moves contracts and supplier prices into private accounts nobody sees. The answer is an approved route that works better because it has company context.
How much can AI decide in procurement without approval?
Only what is cheap and easy to reverse, and only after evidence. Autonomy is a dial per process step, not one policy for everything. Spend classification and deadline monitoring run with spot checks, a letter to a supplier in autonomous mode with approval, and awards and terminations are decided by people. Promotion is earned with data: in our worked example, 6 of 410 index calculations were overridden, 1.5%, after which the calculation ran in autonomous mode with a 10% sample. Coverage grows without control shrinking.
What happens to our procurement data if we switch AI models?
Nothing, if memory and model are kept apart, and that is what you should insist on. Models are converging and available to every competitor. Your advantage lies in contracts, prices, negotiation history and the reasoning behind each decision. In Elevatry these sit in the Company Brain in the EU data zone, separate from the model; the Model Gateway picks the model per task. Tying your intelligence to one model is single-sourcing a critical part. Rent the model, own the mandate.

Friday notes on the topic

Next step

See it on your own procurement.

We show you on one of your processes how Companions do the work from analysis to contract. With your data and your approval rules, if you like.

See the platform