Home · Business · IT, Data & Security · Automation & integrations
As of 13 August 2026, AI can merge data from multiple spreadsheets.
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 ita colleague
What the alternative costsA purpose-built spreadsheet product such as Rows provides built-in AI analysis and live data connections for spreadsheet work.
If this goes wrong, rows can be duplicated, omitted or matched to the wrong person, making the resulting report unreliable.
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 a working folder and copy in the original spreadsheets without editing them.
- Inspect the headers and record the intended matching column, such as customer ID, order number or email address, plus the rows and columns you need in the final file.
- Remove passwords or access restrictions from copies only, then upload the copies and paste the prompt into an AI tool that can work with spreadsheet files.
- Fill in the bracketed requirements in the prompt and ask the tool to create the merged workbook and its audit report.
- Compare the audit report's source and output row counts, unmatched records, duplicate keys and conflict list with the original spreadsheets.
- Open the merged workbook with a colleague, check samples from each source sheet and reconcile important totals before saving a versioned final copy or sending it to the report owner.
Prompt
Merge the attached spreadsheets into one output workbook. Before changing anything: 1. List each file and sheet, including its row count, column names and apparent data types. 2. Identify possible matching keys and explain which columns you would use to match records. 3. Do not invent a matching rule. If a match is ambiguous, keep the record unmatched and list it in an exceptions sheet. Then create the merged output using these requirements: - Source files: [file names or descriptions] - Desired output: [one row per customer, transaction, case or other unit] - Matching key: [column or combination of columns, if known] - Columns to retain: [list, or say retain all] - Conflict rule when values differ: [rule, or flag the conflict] - Duplicate rule: [rule, or retain and flag duplicates] - Required output format: [XLSX or CSV] Preserve the original values and source file or sheet for every output row. Do not silently drop rows, overwrite conflicting values or guess missing data. Produce the merged file if you can. Also produce an audit report containing source row counts, output row count, matched rows, unmatched rows, duplicate keys, conflicting values and any assumptions. If you cannot return a file, provide reproducible Python or spreadsheet steps that perform the merge and explain where I should insert the file names. Finish with a short list of checks I must carry out before using the result.
Open it prefilled in ChatGPT or Claude, or copy it into Gemini, which takes no prefill link.
What it gets wrong
- AI cannot decide whether two similar names or addresses refer to the same real-world record without a rule you can defend.
- AI can produce a plausible merge that drops rows or duplicates them unless the audit counts and exceptions are checked.
- AI cannot know which source is authoritative when two spreadsheets contain conflicting values unless you specify the rule.
- AI cannot take responsibility for reports, customer contact or operational decisions based on a wrong merge.
Even on a YES, the friction has a name: judgement under ambiguity, verification cost and stakes of error.
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 AI combine Excel files into one spreadsheet?
- Yes. It can inspect multiple files, combine compatible tables and create formulas, code or an output workbook. You still need to check row counts, duplicates, unmatched records and important totals.
- Can AI match rows from different spreadsheets?
- Yes, when you provide a clear key such as an ID or a defined combination of columns. It should flag ambiguous matches rather than guessing, because similar names or addresses are not reliable proof of identity.
- Will AI preserve all my spreadsheet data when merging?
- Not automatically. Ask for source tracking, an exceptions list and an audit report, then compare the output with every source file to find dropped, duplicated or changed records.
- Is it safe to use AI to merge work spreadsheets?
- It can be suitable for a controlled copy of the data, but access controls and your organisation's data policy still apply. Keep the originals, avoid uploading confidential data to an unauthorised service and have a colleague check the result before it drives a report or decision.
Nearby answers
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.