The AI quotes an old price. What next?

Sales receives an enquiry quoting a price discontinued months ago. The prospect copied it from an AI answer. The conversation now begins with an avoidable correction.

A general request to update the model will not resolve that specific finding. Investigate the claim and its possible sources instead.

Compare the full offer

Save the answer with its question, provider and date. Compare more than the amount: currency, contract duration, billing unit and tax treatment matter. Does it concern the current offer or an older plan?

An apparent pricing error may be a missing qualification. That can still mislead a customer, but the correction needs to identify what was omitted.

Find where the old figure remains

Open the supplied sources. Check the company pricing page, structured data, older PDFs and commonly linked comparisons as well. Not every wrong answer links to evidence establishing where the claim originated. Keep findings separate from guesses.

For company content, a responsible team should reconcile the relevant locations. A historical label may be more appropriate for an old price list than silently changing the document. An error on another website can be supported with the current original source.

Observe after correcting

Ask the same question later under comparable conditions. Does the old price remain? Is a different source cited? Are qualifications preserved?

Record the result as another observation. One correct answer does not establish that the error has disappeared everywhere. A remaining error does not establish that the correction was pointless.

For recurring cases, maintain a short record of the price version, source locations and changes. The next sales enquiry can then start from what the team already knows.