AI & client relationships
Should I Tell Clients I'm Using AI?
When disclosure builds trust, when it backfires, and how to write a policy that protects the relationship.
Book a discovery callThe short answer
Usually yes - but how you disclose matters as much as whether you do. Proactive, confident disclosure in the right context builds trust. Defensive, apologetic disclosure - or granular detail nobody asked for - creates doubt that didn't exist. And in some situations, disclosure is contractually required regardless of preference.
The question most service businesses are actually wrestling with isn't whether to tell clients - it's how to do it in a way that protects the relationship rather than introducing unnecessary anxiety. The answer depends on your sector, the nature of the work, your client relationship, and increasingly, the terms of your client's contract.
This guide covers the four situations where disclosure clearly builds trust, the five situations where it tends to backfire, and sample wording you can use directly in contracts, proposals, and client conversations.
When disclosure builds trust
Four situations where telling clients works in your favour
In these scenarios, proactive disclosure is consistently the stronger commercial and relationship choice. The common thread: the client already has some expectation that you'll use the best available tools, and confirmation of that - framed confidently - reinforces rather than undermines their trust in you.
You're in a relationship-based business where trust is the product
Coaches, consultants, therapists, advisers - the client is paying for your thinking and judgment, not a generic output. In these relationships, proactive transparency about how you work builds rather than erodes trust, because it demonstrates confidence in the quality of what you deliver regardless of the tools used.
How to do it
Disclose early and proactively, in the context of how your process works. Tie it to the quality of what you produce, not to the existence of the tool.
Example wording
"I use AI to help me do initial research and structure documents quickly - it means I can spend more of our time together on the analysis that's specific to your situation, rather than on drafting mechanics."
The client has asked, or their contract requires it
An increasing number of client contracts - especially from larger organisations, public sector bodies, and professionally regulated organisations - include clauses about AI tool use, data handling, and output verification. If a client asks directly, or their contract requires disclosure, you have an obligation to be truthful.
How to do it
Answer honestly and specifically. Don't conflate 'AI helped me structure this' with 'AI wrote this'. Be clear about what you did versus what the tool did.
Example wording
"Yes - I used Claude to help with the first draft structure and to flag any gaps in the research. I reviewed and rewrote significant sections, and everything in the final document reflects my analysis and recommendations."
You're positioning AI as part of your service offer
If your efficiency, turnaround time, or pricing is partly enabled by AI, telling clients this is a feature, not a confession. Framing it as part of how you deliver good work at a sensible price is a competitive differentiator in 2026.
How to do it
Include it in your proposal or engagement letter as a point of pride, not a caveat. Connect it directly to the client benefit.
Example wording
"Part of how we keep our rates competitive while delivering thorough work is using AI tools for the research and drafting stages - freeing our team to focus on the insight and recommendations that require real expertise."
You're in a regulated sector with specific transparency obligations
Legal practices, FCA-regulated firms, NHS suppliers, and some public sector contractors are subject to guidance from their regulatory bodies on AI tool use. In these contexts, voluntary disclosure ahead of being asked is lower risk than reactive disclosure.
How to do it
Review your regulatory body's current AI guidance. Include a brief AI tool disclosure in your standard engagement terms specific to the relevant obligation.
When it backfires
Five ways AI disclosure goes wrong
These aren't arguments against disclosing - they're arguments against disclosing badly. Each scenario describes a common mistake and a better alternative.
Disclosing defensively, with an apologetic tone
"I should mention that I used AI for parts of this" - said nervously, or unprompted, in a way that invites the client to scrutinise the work more critically - plants a doubt that wasn't there. If your tone signals that you think the use of AI reduces the value of what you've delivered, the client will absorb that view.
Do this instead
If you're going to disclose, be matter-of-fact about it. "I used AI for the initial research pass" said confidently is very different from "I should probably mention that AI was involved" said apologetically.
Disclosing granular tool detail nobody asked for
"I used Claude 3.5 Sonnet for the first draft, then ran it through Perplexity for fact-checking, and used Grammarly to polish the final version" - this level of detail focuses the client's attention on your process rather than your output. Most clients don't need or want this. It creates questions that don't need to be asked.
Do this instead
Disclose the existence and role of AI tools in your work, not the specific stack. The client cares whether you verified the output and applied your expertise - not which model you used.
Disclosing before the work has been reviewed by the client
Telling someone 'this was written with AI assistance' before they've read it anchors their evaluation negatively. They read looking for AI fingerprints rather than reading for quality. If the work is good, let them read it first.
Do this instead
If disclosure is warranted, a brief note after delivery - or as part of the cover note explaining your process - is better than a pre-emptive disclaimer that colours their first read.
Signing a no-AI contract clause and then using it anyway
The practical risk of being caught is real - AI-detection tools are used by some clients, and stylistic patterns in writing can be noticed. More importantly, the relationship damage from discovery is significantly worse than either complying with the restriction or negotiating it before signing.
Do this instead
If a client's standard contract restricts AI use and you rely on it professionally, negotiate the clause before you sign. Most clients are open to adding verification and accuracy commitments in exchange for permitting AI tools.
Treating 'AI-assisted' and 'AI-generated' as the same thing
Using AI to help structure a document you then wrote, reviewed, and took responsibility for is fundamentally different from pasting an AI output into an email without editing it. Conflating these in a disclosure - in either direction - misrepresents what actually happened.
Do this instead
Be specific about your role: what did you write, review, verify, and take responsibility for? That specificity is what the client actually needs to know.
Sample wording
Ready-to-use language for five situations
These are starting points, not final versions - adapt to your voice and sector. The contract clause in particular should be reviewed against your specific service type and any sector-specific regulations that apply to you.
Add to your standard engagement letter or service agreement
We may use AI tools to assist with research, drafting, analysis, and document structuring as part of our standard workflow. All work is reviewed, verified, and approved by a qualified member of our team before delivery. You retain full ownership of all deliverables as set out in this agreement.
Short, clear, and covers the three things clients actually care about: that AI is used, that a human reviewed it, and that their ownership rights aren't affected.
Add to the methodology section of a proposal
Our process uses AI tools to accelerate research and first-draft work, allowing us to focus our time on the analysis, strategy, and recommendations that require direct expertise. All client deliverables are reviewed and approved by the lead consultant before submission.
Frames AI as a process efficiency that benefits the client - more senior time on the work that matters.
Use when proactively explaining your process to a client
I wanted to give you a quick note on how we work. We use AI tools - primarily for research gathering and initial structuring - as part of our standard process. Everything you receive from us has been reviewed, edited, and signed off by me personally. If you have any questions about our process or have specific requirements around AI tool use, please let me know and we can discuss.
Personal, warm, and invites a conversation rather than closing it down.
Use if a client asks whether you used AI on a specific piece of work
Yes - I used [tool] to help with the initial research and structure, which I then reviewed thoroughly, rewrote in sections, and built on with my own analysis. The final document represents my recommendations and judgment. Happy to walk you through my thinking on any section if that would be useful.
Honest, specific about your role, and immediately redirects to the quality of the thinking rather than the tool.
Use when negotiating before signing
I'd like to discuss clause [X] around AI tool use before we proceed. Our process does involve AI tools for research and drafting efficiency, and I'd want to be transparent about that rather than agreeing to a restriction I couldn't guarantee I'd always meet. I'm happy to add stronger verification and accuracy commitments in exchange - would that work for you?
Professional, pre-emptive, and offers a trade. Far better than signing and hoping.
Need an AI policy for your business?
We help service businesses build AI workflows and the policies that go with them
From choosing the right tools to drafting client-ready disclosure language and internal use policies, our consultancy sessions cover the practical and relational side of AI adoption together.
Common questions
FAQs
Get in touch