Home · Business · Admin & Office · Procurement
As of 13 August 2026, AI can check a supplier's reputation.
This still needs a person who signs their name to it.
Can you do it?
15 minutesto a draft.
30 minutesto something you’d act on.
Cost, all in£0
Skill neededchat-fluent
Who has to check ityou
What the alternative costsNo human-service price is supplied in the available tool data.
If this goes wrong: you select a supplier whose problems were missed or whose public claims were overstated, causing delays, wasted spend or a failed contract.
What to actually do
Hand it to a person
The route this page recommends
A person who owns the outcome does this end to end, worth it when the failure is dear.
Use a tool built for this
Second choiceDo it yourself
The distant thirdA chat interface, chat-fluent skill, and roughly 30 minutes until you can act on the result.
How to actually do it
- Open the supplier's website, any quotation or tender documents, and the relevant Companies House or official-register pages, then record the exact legal name, company number, trading address and claimed accreditations.
- Gather the supplier's website URL, sector, proposed purchase, contract importance, selection requirements and any specific concerns, and paste these into the prompt.
- Paste the completed prompt into a web-enabled chatbot and ask it to use direct public sources, cite each material claim and separate facts from reviews, allegations and inferences.
- Open every source link in the brief, compare the legal identity and dates with the supplier's own documents, and mark each important claim as confirmed, conflicting or unverified.
- Send the unresolved questions and any serious warning signs to the supplier, request documentary evidence where appropriate, and record its written replies.
- Compare the verified evidence and replies against your procurement requirements, then have the authorised colleague approve the supplier decision and retain the evidence with the procurement record.
Prompt
Check the reputation of this UK supplier using current, publicly available sources that you can access. Supplier name: [SUPPLIER NAME] Supplier website: [WEBSITE] Sector and goods or services: [SECTOR] Proposed purchase and approximate importance to the business: [PURCHASE] Requirements that matter: [REQUIREMENTS] Specific concerns to investigate: [CONCERNS] Produce a procurement due-diligence brief dated 2026-08-13. Separate verified facts, customer reviews, allegations, company claims and your own inferences. Check the supplier's legal identity, trading history where available, contact details, relevant accreditations, financial or operational warning signs, complaints and review patterns, notable news, and evidence of comparable customers or work. Use UK sources where relevant, including Companies House and official registers. Cite every material claim with a direct source link and explain the publication date or access date where available. Do not invent facts, fill gaps with assumptions, treat review volume as proof of quality, or present a lack of search results as proof that no problem exists. Flag conflicting sources, unverifiable claims, possible duplicate or manipulated reviews, and information that is old or incomplete. End with a table of strengths, risks, evidence quality and unanswered questions, followed by a cautious recommendation of proceed, seek more evidence or do not proceed. Do not give legal advice and do not make the final procurement decision for me.
Open it prefilled in ChatGPT or Claude, or copy it into Gemini, which takes no prefill link.
What it gets wrong
- AI cannot prove that an anonymous review, complaint or allegation is genuine or representative.
- AI cannot replace a reference call, site visit, sample inspection or other physical check of how the supplier operates.
- AI cannot know which level of supplier risk your organisation is prepared to accept.
- AI can miss recent, unindexed or private information and can overstate a conclusion when the public evidence is thin.
Even on a YES, the friction has a name: real time truth, verification cost and judgement under ambiguity.
How we scored this
Five axes, each scored nought to two by hand: ten means AI carries the task cleanly, and the thresholds that turn a total into YES, PARTLY or NO are published in the methodology. Each axis name links to its definition.
| Axis | Score (0–2) |
|---|---|
| Output | 2 |
| Inputs | 2 |
| Verification | 1 |
| Liability | 1 |
| Effort delta | 2 |
| Total | 8 / 10 |
The methodology and its thresholds are published in full.
FAQ
- Can ChatGPT check whether a supplier is trustworthy?
- Yes, it can assemble a dated public-evidence brief covering company records, reviews, news and stated accreditations. It cannot prove that every review or allegation is genuine, so you must open the sources and confirm important claims before selecting the supplier.
- What should AI look for when checking a supplier?
- Ask it to check the supplier's legal identity, trading history, official registrations, accreditations, complaints, review patterns, news and evidence of comparable work. It should separate verified facts from company claims, allegations and inferences.
- Can AI tell me which supplier to choose?
- It can compare the evidence against requirements and identify risks or unanswered questions. The authorised person in your organisation must decide whether the risk is acceptable and whether further checks, references or contract protections are needed.
- How do I verify an AI supplier reputation check?
- Open every cited source, confirm that it refers to the correct legal entity, and check the dates and wording against the brief. Ask the supplier about unresolved points and retain the evidence and replies with the procurement record.
Nearby answers
- Can AI compare software suppliers for my business?PARTLY
- Can AI find cheaper supplies for my business?PARTLY
- Can AI manage stock reorders for my business?PARTLY
- Can AI source branded merchandise for my business?PARTLY
- Can AI track my business orders?PARTLY
- Can AI chase a supplier about a late delivery?YES
Assessed by gpt-5.6-luna (gpt-5.6-luna) on 2026-08-13, second-checked by an independent model. Wrong somewhere? Email [email protected] and it gets re-checked.
The newsletter
AI news, new answers and product picks, straight to your inbox.