How Architects Are Really Using AI: The Data and Three Composite Stories
If you feel behind on AI, you probably aren't.
In March 2025, the American Institute of Architects published research on AI in architecture firms (AIA press release). Here's what it found:
- Only 6 percent of architects use AI regularly in their work.
- Only 8 percent of firms have implemented AI solutions. Another 20 percent are working on it.
- 84 percent of respondents were optimistic about AI automating manual tasks, and 74 percent about product research.
- 90 percent had concerns about risks like inaccurate results, security, and transparency.
So most architects are curious, careful, and early. That's a healthy place to start.
Three composite stories
The stories below are composite examples, not real people. Each one combines patterns that are common when architects use AI well. We don't attach numbers to them, because made-up numbers wouldn't help you. Your own results are the ones that count.
The solo practitioner: less admin, more design
The situation: A solo architect runs several small residential projects at once. Emails, meeting notes, and code research fill the days.
What they use AI for:
- A first pass on code research for renovations, with official sources to check
- Drafting client updates in plain language
- Turning site meeting notes into action items
What changes: The blank page disappears. Each task starts from a draft instead of from nothing. The architect still reads every code section and edits every email, but the work goes faster, and more of the week goes to design.
The project manager: clearer coordination
The situation: A project manager at a small firm coordinates a healthcare project with many consultants and strict regulations.
What they use AI for:
- Organizing regulatory research into checklists by topic and consultant
- Summarizing long consultant emails into decisions and open questions
- Explaining technical issues to the client in simple terms
What changes: Fewer loose ends. Every meeting ends with a clear list of who does what. The client understands why a structural review or a code question matters before it becomes a surprise.
The young architect: more options, explored faster
The situation: A young architect wants to explore more design ideas within tight project schedules.
What they use AI for:
- Precedent research to start design conversations
- Mood images to show a client the feeling of a space early
- Testing whether a design narrative is clear to a non-architect
What changes: More alternatives get explored before the team commits. The architect knows AI images are not construction information, and presents them only as early ideas.
What the three stories have in common
They start with one painful task. Not "AI for everything." One task, every day, for a week.
They keep judgment in human hands. AI drafts, summarizes, and organizes. The architect decides.
They verify. Every code answer gets checked against the official source. Every client email gets read before it's sent.
They protect confidential information. They remove client names before pasting notes into a free tool, or they use a business plan with stronger privacy controls.
How to measure your own results
Skip the industry averages and track your own:
- Time: How long does a task take with AI, compared with before? Try it on the same kind of task for two weeks.
- Quality: Are your emails clearer? Are fewer action items missed?
- Client response: Do clients ask fewer follow-up questions?
- Comfort: Do you trust the workflow more each week, or less?
If a task isn't getting faster or better after two weeks, drop it and try another.
Where to start
Pick one task you dislike: meeting notes, a zoning PDF, or a client email that's hard to word. Use AI on it every day for a week, and save the prompts that work.
Since only 6 percent of architects use AI regularly, a few weeks of steady practice puts you ahead of most of the profession, while keeping the careful habits that make you a good architect.