George Proud, CEO of Gendo, answers readers’ questions on how AI is transforming architectural visualisation, opening up creative opportunities for practices of all sizes while keeping architects in control of the design.

Visualisation of Settle’s Frame agrotourism retreat, developed in Gendo, with co-founders Ed and Andrew collaborating live on the same board.

How is AI changing the way architects approach visualisation?
The biggest change is that visualisation is becoming part of the design process, rather than something that happens at the end. Producing high-quality architectural imagery used to require time, specialist software and technical expertise. Today, an architect can turn a sketch, a 3D model or a screenshot into a convincing visual in seconds.

The real opportunity isn’t just speed, though. When something that took hours takes minutes, you can explore more ideas and involve clients much earlier. In The AI Workflow Shift in Architecture, our 2026 user research based on over 1,000 responses, 78% of respondents answering the question reported saving time in the previous month using Gendo. Visualisation is becoming a tool for developing ideas, not just communicating finished ones.

Can AI help smaller practices compete with larger studios that have dedicated visualisation teams?
Yes, and I’d argue smaller practices stand to gain the most. Smaller studios rarely have dedicated visualisation teams. They produce renders themselves, alongside everything else, or outsource the work, which can be expensive and slow. In our recent survey, nearly half of respondents reported reducing their use of traditional rendering software, while 19% had reduced their reliance on external CGI studios or freelancers. We’ve seen this with Settle, a two-person London studio that reduced its visualisation time per image by around 80% using Gendo. AI has the potential to level the playing field, letting smaller practices compete on the quality of their ideas rather than the size of their production teams.

Where in the design process can AI visualisation add the most value?
I’d say early-stage design, when ideas are still developing and there’s the greatest opportunity to influence the outcome. An architect can start with a SketchUp model or hand-drawn concept and explore different materials, lighting or architectural approaches in a single working session, rather than over days.

Among survey respondents, 55% use Gendo for design development and exploring options, while 58% use it for client presentations and design reviews. That shows visualisation becoming part of the whole design process, not just the final presentation. Client conversations benefit too. If a client asks how a building would look with a different façade material, you should be able to explore it together, not arrange another meeting next week.

How can architects use AI to explore ideas without losing control of their design intent?
Asking AI to invent a building is very different from asking it to visualise one you’ve already designed. Architects need confidence that their geometry, proportions and materials are being respected. If a tool adds a floor or changes the windows, it may produce a beautiful image, but it isn’t your design. That’s why, rather than simply generating images from text, we’re building tools that work with an architect’s existing design information and allow targeted changes to specific elements. The architect should always be the author of the design. AI should expand their ability to experiment, not take creative control away.

What are the risks and limitations architects should be aware of when using AI-generated imagery?
The biggest risk is confusing a convincing image with an accurate one. A realistic visual doesn’t mean the materials are appropriate, the geometry is correct or the building could be constructed as shown. That matters enormously for planning applications, client approvals and investment decisions, so professional judgement applies whether or not AI was involved.

There are also considerations around intellectual property, confidentiality and data protection. Practices should understand where their project information is stored and whether it’s being used to train AI models. AI should also create opportunities for younger architects, not remove the experiences through which they build essential skills. Ultimately, AI doesn’t replace architectural knowledge, responsibility or judgement.

What advice would you give practices looking to integrate AI visualisation into their workflow for the first time?
Start using it, and start with a real project. Larger organisations especially can spend months evaluating platforms and drafting policies before anyone has experimented. Those conversations matter, but so does practical experience. Take an existing sketch or model, bring AI into a familiar workflow and measure what changes.

Make it accessible across the team rather than a specialist tool only one person can operate. And don’t treat it purely as cost-saving. The bigger opportunity is doing what previously wasn’t practical: exploring more alternatives, responding to feedback faster and giving clients a greater role in the creative process. The practices that benefit most won’t have the biggest budgets. They’ll be the ones that experiment, learn and integrate AI into their creative process, while keeping architects firmly in control of the design.

For more information please visit gendo.ai