Last updated: June 2026.
Can ChatGPT format candidate CVs for a recruitment agency? Yes — for one-off documents it does a reasonable job. At agency scale it breaks: no fixed branded templates, no batch processing, inconsistent output between runs, no ATS integration, no audit trail, and genuine GDPR questions about pasting candidate personal data into a consumer chatbot. Purpose-built tools format a CV in about 60 seconds for around $0.99.
This post is the honest version of that answer: where ChatGPT genuinely works, exactly where it falls over, and how agencies are combining AI assistants with formatting engines instead of choosing between them.
Can ChatGPT format a candidate CV at all?
Yes, within limits. Paste a candidate's CV text into ChatGPT with a clear prompt — "restructure this into reverse-chronological order with sections for profile, experience, education and skills" — and you'll usually get well-organised text back. It can tighten wording, standardise tense, summarise a career history and reorder content. For a single CV where the output destination is plain text, that's genuinely useful.
What ChatGPT returns, though, is text — not a document. It cannot reliably produce your agency's branded DOCX template with your logo, fonts, margins, header and footer. You still have to copy the output into Word and rebuild the layout manually, which is most of the work you were trying to avoid. Manual CV formatting takes 20-40 minutes; ChatGPT typically removes five of those minutes and leaves the rest.
Where does ChatGPT break at agency scale?
Five places, consistently:
- No fixed branded templates. Every client template carries precise branding rules. ChatGPT outputs prose, not pixel-faithful documents — you cannot lock a layout and have every CV come out identical.
- No batch processing. Twenty CVs for a Monday shortlist means twenty separate conversations, twenty copy-paste cycles, twenty manual rebuilds in Word.
- Inconsistent output between runs. The same prompt on two CVs — or the same CV twice — produces different section orders, headings and phrasings. Clients notice when a shortlist isn't uniform.
- No ATS integration. Nothing flows back into Bullhorn, JobAdder or Vincere; every document is a manual round trip.
- No audit trail. When a client or regulator asks who processed a candidate's data, when, and what was changed, a chat history spread across personal accounts is not an answer.
None of these are model-quality problems. GPT-5-class models write excellent prose; they simply aren't document-production systems.
Is it safe to paste candidate CVs into ChatGPT?
This is the question agency owners should ask first. A CV is a dense bundle of personal data — name, contact details, employment history, sometimes date of birth or photo. Under UK and EU GDPR, your agency is the data controller, and pasting that data into a consumer chatbot means transferring it to a processor without a signed data processing agreement (DPA), without certainty about training use on free tiers, and without control over retention or storage location.
Consumer ChatGPT accounts don't come with a DPA; business tiers improve the position but still leave templates, batch and audit gaps. A recruiter quietly using a personal ChatGPT account to format CVs is shadow IT handling special-category-adjacent data. Purpose-built CV tools are designed for this: FormaCV, for instance, signs DPAs, offers EU data residency, and keeps an audit log on every plan. Whatever tool you choose, the data-processing question is not optional.
When is ChatGPT genuinely fine?
Be fair to the chatbot — there are recruitment tasks where it's the right tool:
- Drafting text, not documents: candidate summaries, ad copy, outreach emails, interview question lists.
- One-off internal formatting where no client branding is involved and the candidate's data has been stripped or the candidate is yourself.
- Brainstorming structure for an unusual CV — a portfolio career, an academic-to-industry switch — before a human writes it.
- Low-volume desks placing a handful of candidates a month, where 20 minutes of rebuild per CV is tolerable and the data risk is consciously accepted with candidate consent.
If that describes your desk, ChatGPT plus a Word template may be all you need. The breakage starts when volume, consistency, or compliance enters the picture — usually around the point an agency submits more than a few dozen CVs a month.
What do purpose-built CV formatting tools add?
Purpose-built AI CV formatting tools take the same underlying AI capability and wrap it in document production: fixed branded templates that render identically every time, batch upload, anonymisation for blind CVs, ATS integration, translation, and an audit trail. FormaCV formats a CV into a locked branded template in about 60 seconds for $0.99 per CV, with unlimited templates per client or branch, GDPR anonymisation, 40+ language translation, and unlimited AI tailoring included — and a 30-day free trial with no credit card. Comparable tools (HireAra, Allsorter, Candidately, iReformat) make the same structural trade: less conversational flexibility than ChatGPT, far more consistency and compliance. Our 8-tool comparison walks through which fits which agency. The point isn't that these tools have better AI than OpenAI — it's that they put the AI inside a controlled document pipeline.
Can you use ChatGPT-style AI and a formatting engine together?
Yes — and this is where the market is actually heading. The Model Context Protocol (MCP) lets AI assistants like Claude, Cursor, and other MCP clients drive external tools directly. FormaCV ships a native MCP server (open-sourced at github.com/Rocketech-Software-Development/formacv-mcp), so a recruiter can stay in a chat interface — "format these five CVs into the Acme template and anonymise them" — while the actual document production happens inside FormaCV: locked templates, audit log, DPA-covered processing.
That's the resolution to the ChatGPT question. The conversational interface is genuinely good for recruiters; the consumer chatbot as a data processor and document engine is not. Splitting the two — AI assistant for intent, formatting engine for execution — gives you the ergonomics of chat without rebuilding documents by hand or routing candidate data through an unaudited tool.
The bottom line
ChatGPT can format a CV; it cannot run your agency's CV formatting. Use it freely for drafting and one-offs where no candidate PII is at stake. The moment branded consistency, batch volume, ATS flow or GDPR accountability matters, use a purpose-built engine — and if your team loves the chat workflow, connect the two through MCP rather than pasting candidate data into a consumer app.