A 250-Person Firm's One-Page AI Use Statement For RFPs

From Mission Statement to AI Use Statement

An AI use statement is a short, client-facing statement of how your firm uses artificial intelligence responsibly— which tasks AI assists, how licensed professionals review the work, and how the firm keeps accountability and confidentiality intact. Where a mission statement is aspirational, an AI use statement is operational.

Firms have always articulated who they are in public. The mission statement, the values page, the design philosophy on the About section: these say what the practice stands for and why clients should trust it. That instinct is sound. The subject just expanded.

Clients now want to know not only what you believe, but how you work in an age when a tool can draft a spec section or render an option in seconds. A mission statement alone no longer covers that. An architecture firm mission statement was written to express purpose, not to answer "did AI touch our drawings, and who's accountable if it did?"

The AI use statement answers exactly that. It's the operational companion to the mission statement: the mission says what the firm stands for; the AI use statement says how it works responsibly with AI— and a licensed professional stays accountable for every deliverable.

If it's so useful, why doesn't your firm already have one? Until recently, nothing forced the question. That changed.

Why Your Firm Needs One Now

Most firms now use AI every day and say nothing about it. About 60% of architecture practices report using AI, up from 41% a year earlier, yet only 15% have a formal AI policy in place. Those figures come from the RIBA AI Report 2025¹ (UK data), but the pattern travels. Three forces are closing that gap fast.

The adoption-versus-policy gap. The same RIBA research¹ found that over half of practices expect to develop an AI policy within two years. Most firms are already past the question of whether they use AI. What they haven't done is state a position— which means the first time a client asks, they're improvising. Building a firm-wide AI governance practice starts with writing down what's already true.

The profession is asking for transparency. The AIA Trust advises architects to be upfront with clients about generative-AI use and to explain how those tools are responsibly incorporated². Best practice is to do it early, before a project begins, while clients can still raise concerns. The AIA's own guidance frames AI as supporting, not replacing, professional judgment³. Transparency is where the profession's guidance already points.

RFPs and contracts are starting to require it. The clearest signal is federal. The GSA's draft procurement clause (GSAR 552.239-7001) would require contractors to disclose all AI systems used in contract performance within 30 days of award⁴. It's a draft, it's federal, and it's been deferred, so treat it as the leading edge, not the rule today. Read alongside the AIA's transparency guidance and the liability questions AE firms already field, the direction of travel is clear.

None of this requires a fifty-page framework. It requires one page that sits on top of an AI strategy that goes beyond a single paragraph: consistent review standards and data handling your team actually follows.

It also doesn't mean announcing every spell-check. The real question is where the line sits.

When You Have to Disclose AI Use (and When You Don't)

Disclose AI use when AI materially shapes a client deliverable— especially when AI-generated content reaches the client without human review. Internal, human-reviewed tool use generally doesn't require disclosure. The dividing line is professional review.

That line has an ethical spine. Under the AIA Code of Ethics (Rule 4.102), architects must maintain "responsible control" of their work— defined as "the degree of knowledge and supervision ordinarily required by the professional standard of care"². Responsible control is the standard that decides how much AI involvement triggers review, and in turn disclosure. If a person with a license reviewed it and stands behind it, you're inside the standard. If AI output reached the client untouched, you're not.

Legal commentary on professional-services AI disclosure draws the same line: disclosure is warranted when AI-generated reports or analyses reach clients without human review, and is generally optional when AI is an internal aid under professional oversight⁵. Risk guidance for AE firms puts it plainly— if AI figures prominently in your work, you likely will want to disclose it⁶.

Disclose when:

• AI-generated content reaches the client without human review
• AI interacts with the client directly
• AI drives a significant project decision
• AI processes client personal or confidential data

Generally don't need to when:

• AI helps with internal research, first drafts, or summarization— and a licensed professional reviews and approves every output before it leaves the building

One caveat: where "material" begins is a judgment call. Some firms build a tiered disclosure model (none, light, or full) scaled to how much AI shaped the work. Set your own threshold with leadership and counsel. This is a framework, not legal advice.

Once you know where the line is, you can write the document. Here's what belongs on the page.

What Goes In a One-Page AI Use Statement

A one-page AI use statement covers five things: the scope of AI use, the human-review standard, data and confidentiality handling, an ownership and IP stance, and a named point of accountability. Keep it to a single page in plain language. Each component maps to one or two sentences you'll actually write.

1. Scope of use. Name where AI is and isn't used. Research, drafting, analysis, and quality checks are common yes-zones; autonomous design or engineering decisions are the clearest no. Specific beats vague. The reader wants to know what AI did, in concrete terms.

1. Human-review / responsible-control standard. State that a licensed professional reviews and is accountable for every deliverable. This is the heart of the statement and exactly what the AIA's "responsible control" standard asks for². It's also the line that turns disclosure from confession into reassurance³.

1. Data and confidentiality. Say what client and project data does and doesn't go into AI tools. Risk guidance for AE firms is blunt here: don't feed sensitive or confidential material to tools that retain inputs to train their models⁶. Name how you keep project data out of them.

1. IP and ownership. Take a stance on who owns AI-assisted work product and inputs. This is moving under your feet— the GSA's draft clause would have the government keep ownership of AI inputs and outputs, with the contractor getting only a limited license⁴. Some clients and clauses will claim rights, so coordinate the specifics with counsel.

1. Accountability / point of contact. Name a role responsible for the firm's AI practices and for answering questions on a given project. Accountability that isn't assigned isn't accountability.

Components are the ingredients. Here's the page itself.

A One-Page AI Use Statement Template

Here is a fill-in skeleton you can adapt today. Replace the bracketed prompts with your firm's specifics, keep it to one page, and run the final language past your counsel and professional-liability (PI/E&O) insurer before it goes into a contract.

[Firm Name] AI Use Statement. We use AI tools to help our team work faster and more thoroughly— research, drafting, analysis, and quality checks. We do not use AI to make professional design or engineering decisions. A licensed [architect/engineer] reviews and is responsible for every deliverable we issue. We protect client and project information and do not put confidential data into tools that would use it to train their models. We retain professional responsibility for our work product under [applicable standard]; questions about AI on a specific project go to [name, role, contact]. Last reviewed: [date].

Read it back. It's plain, it's human-led, and it fits on one page. Write it as AI disclosure that doesn't read like a legal footer, because the people reading it are deciding whether to trust your firm. They aren't auditing your compliance.

Two practical notes. Review the statement at least annually, more often if your tools or your clients' expectations shift faster than that. And keep the "coordinate with counsel and your insurer" habit attached to anything that lands in a contract; this template speaks in your firm's voice, and the contract language is counsel's job.

The same statement does different jobs in different places. Here's how it flexes.

How the Statement Flexes— RFP, Website, and Contract

One statement, three venues. In an RFP it's a confident paragraph that answers the AI question directly. On your website it's a short public commitment. In a contract it's precise language you draft with counsel.

In a proposal, your AI use statement answers the procurement officer and reassures the design-led client in the same breath— human-led, reviewed, accountable. Lead with the review-and-accountability line, answer the literal question, and project confidence instead of defensiveness. This is the moment the title promised: the RFP asks, and you have a real answer ready.

There's a quiet fear under this— that admitting AI use makes a design-led client feel the work is less yours. Framed as human-led, it does the opposite. You're telling a sophisticated client that a licensed professional stands behind every page, and that you're transparent enough to say so. That builds trust.

The contract version is where you stop freelancing. Ownership, data, and right-to-rely language belongs with your counsel and insurer²⁴. If you're adding an AI clause to new contracts, that's a lawyer's pen at work, not a marketing edit. Keep the core promise identical across all three venues so you're never caught saying different things in different places.

A few questions come up every time a firm writes its first one.

FAQ

Short answers to the questions firms ask first.

Do architecture firms need an AI use statement?

Increasingly, yes. Clients, RFPs, and professional-ethics guidance are starting to expect a stated position on AI use, and most firms already use AI without one²⁴. The RIBA AI Report 2025 (UK) found about 60% of practices using AI but only 15% with a formal policy¹. That gap is the need.

What should an AI use statement include?

Five things on one page: the scope of AI use, the human-review standard, data and confidentiality handling, an IP and ownership stance, and a named point of accountability²⁶. Write it in plain language. This is the document a client reads to decide whether to trust you.

When do you have to disclose AI use to a client?

When AI materially shapes a deliverable, especially if AI-generated content reaches the client without human review⁵. Internal, human-reviewed tool use generally doesn't require disclosure. Under the AIA's "responsible control" standard, a licensed professional accountable for the output is the dividing line².

Does the AIA require disclosing AI use?

The AIA recommends transparency and being upfront about generative-AI use; under its Code of Ethics, architects must keep "responsible control" of their work². The AIA frames AI as supporting, not replacing, professional judgment³.

Does disclosing AI hurt us competitively?

Framed as human-led, it builds trust and sets you apart. The risk is poor framing that implies AI replaced professional judgment. In RIBA's UK survey, more than nine in ten architects rejected the idea that AI could substitute for professional decision-making¹. Say it the way the profession already believes it.

The statement is the easy part. Living up to it is the practice.

The Statement Is a Promise

Writing the paragraph takes an afternoon. Building the practice behind it (consistent review standards, data handling your team actually follows, language that holds up across every proposal) is the real work.

An AI use statement is a promise. The firms that win trust are the ones whose practice already makes the promise true before they write it down. That's the whole thesis: AI drafts and analyzes, but it doesn't carry a license or stand behind a stamp. A person does. The AIA puts it the same way— AI supports, not replaces, professional judgment³.

If turning that paragraph into a real governance practice (mapped to your workflows, your tools, and your proposals) sounds like more than an afternoon, that's the kind of work an AI strategy partner can help you scope. Start with the page. Build the practice it promises.

References

1. Royal Institute of British Architects, "RIBA AI Report 2025" (2025) — https://www.riba.org/work/insights-and-resources/ai-report/riba-ai-report-2025/
2. The AIA Trust, "Ethical Challenges of Generative AI in Architectural Practice" (2025) — https://theaiatrust.com/ethical-challenges-of-generative-ai-in-architectural-practice/
3. The American Institute of Architects, "Guidance for the Responsible Use of AI by Architecture and Design Firms / AI Task Force" (2026) — https://www.aia.org/resource-center/ai-task-force
4. Gibson Dunn, "GSA AI Procurement Rules Would Introduce New Disclosure and Use-Rights Requirements for Federal Contractors" (2026) — https://www.gibsondunn.com/gsa-ai-procurement-rules-would-introduce-new-disclosure-and-use-rights-requirements-for-federal-contractors/
5. Fourscore Business Law, "AI Disclosure in Business: When, Why, and How to Inform Clients About Your Use of AI" — https://www.fourscorelaw.com/resources/ai-disclosure-in-businessnbspwhen-why-and-how-to-inform-clients-about-your-use-of-ai
6. B Brown, "Why Architectural and Engineering Firms Need AI Ground Rules" (2023) — https://us.bbrown.com/blog/why-architectural-and-engineering-firms-need-ai-ground-rules