AI Is becoming part of architectural design, but can your firm trust the project information it uses?

AI Is Moving Directly Into Design Workflows

In October 2026, Graphisoft introduced its 2026 product portfolio, including new artificial intelligence capabilities for architecture, engineering and construction teams. The company is developing tools designed to better connect early-stage design, feasibility analysis, project information and BIM environments. This reflects a broader direction across the industry:

AI is gradually moving beyond isolated content creation and toward working directly with the information organizations use to deliver projects.

That makes information management increasingly important.

Consider a Normal Project Workflow

A project begins with an initial design.

The client requests a change.

The architect updates the model.

A new PDF is issued.

An engineering consultant is still reviewing an earlier version.

Another change is approved through email.

The estimator is working from a different document.

A newer version is eventually uploaded to a shared folder.

Everyone is working.

But they may not all be working from the same information.

Without AI, that situation already creates inefficiency.

With AI, information can potentially be retrieved and analyzed much faster.

But speed only creates value when the underlying information is reliable.

Faster Information Retrieval Does Not Fix Version Control

Imagine a project manager asking an AI assistant:

“Summarize the latest client-approved changes to the lobby.”

The AI could potentially review:

• emails;
• meeting notes;
• drawings;
• specifications;
• BIM model information;
• consultant documents.

But which source actually represents the final decision?

Which document is still a draft?

Which version has been superseded?

Which modification was formally approved?

If several versions remain accessible without a clear status, AI can accelerate information retrieval without resolving the organization’s existing ambiguity.

What appears to be an AI problem may actually be an information-management problem.

Professional Judgment Remains a Human Responsibility

The Royal Architectural Institute of Canada recognizes the potential of artificial intelligence across several areas of architectural practice, including research, documentation, project administration and analysis.

It also emphasizes that AI does not replace the architect’s professional judgment, duty of care or accountability.

That leads to a straightforward operating principle:

AI can support the workflow. It should not blur responsibility for professional decisions.

Five Foundations to Establish Before Automating Further

1. One Clearly Identifiable Source for Current Project Information

Employees and consultants should know where officially approved information resides.

That does not necessarily mean everything must exist in one application.

It does mean the firm should be able to clearly distinguish between documents that are:

• in preparation;
• under review;
• issued;
• approved;
• superseded or outdated.

That distinction should not depend solely on an employee’s memory.

2. A Clear Change History

For important changes, the project team should be able to quickly answer:

• What changed?
• When was it changed?
• Who requested the change?
• Why was the change made?
• Who approved it?

The more structured this information becomes, the more useful context AI can potentially provide.

3. A Consistent Consultant Collaboration Process

Architecture projects often involve several organizations:

• architects;
• engineers;
• clients;
• contractors;
• estimators;
• specialized consultants.

They may not all use the same systems.

The workflow should therefore clearly define rules around:

• document exchange;
• version control;
• external sharing;
• permissions;
• approvals;
• document retention.

4. Structured Project Communications

An important design decision should not become difficult to locate simply because it was made in an email thread.

In Québec, the Ordre des architectes du Québec confirms that architects’ project files and archives may be maintained digitally and emphasizes the importance of retaining relevant communications within the appropriate project record.

That discipline becomes even more important when AI assistants are expected to retrieve historical project context.

5. Human Validation Before Professional Use

AI can potentially accelerate:

• information retrieval;
• summary preparation;
• document comparison;
• certain drafting activities;
• administrative coordination;
• report preparation.

But outputs used in professional decisions or deliverables should still receive the appropriate level of human review.

The more deeply AI becomes integrated into project systems, the more clearly the firm should define where human approval remains mandatory.

Start With Operational Problems, Not AI Features

An architecture firm considering broader AI adoption should first determine where its team is actually losing time.

Examples may include:

• searching for the latest approved document;
• preparing meeting summaries;
• comparing different versions;
• distributing approved documents;
• collecting consultant updates;
• preparing repetitive reports;
• organizing project records;
• tracking client approvals.

Then two questions become particularly useful:

Which activities truly require professional judgment?

And which could be partially automated while maintaining human oversight?

That distinction usually leads to simpler, more measurable and more practical automation opportunities.

AI Readiness Is Also an Information-Management Issue

When organizations discuss AI readiness, the conversation often focuses on choosing a platform or purchasing additional licences.

Those decisions matter.

But connected AI also depends on several operational foundations:

• information quality;
• document structure;
• permissions;
• version control;
• workflow consistency;
• clearly defined responsibilities.

Without those foundations, artificial intelligence may simply make existing
operational weaknesses more visible.

With the right structure, it can instead become a practical productivity tool.

How InnoAxis Can Help

InnoAxis Solutions Inc. helps Canadian organizations modernize information and workflows before adding additional layers of automation.

For architecture and engineering firms, that work can include:

• workflow mapping;
• document and information architecture;
• Microsoft 365 or cloud collaboration architecture;
• permissions and external-sharing controls;
• AI use-case assessment;
• secure workflow automation;
• governance controls;
• human-review requirements;
• measurement of operational outcomes.

The objective is not to replace BIM platforms or specialized design applications.

It is to create a stronger digital environment around those systems so information is easier to organize, secure, share and automate.