Photo courtesy the Banff Centre for Arts and Creativity

By Alistair Leyland

I spent a couple days during the middle of March in the shadow of Mount Rundle at the National AI Summit in Banff. There, amidst a collision of tech architects, creative leaders and policymakers, the conversation wasn’t just about algorithms. It was about the future of Canada’s creative economy, a sector that contributes more than $60 billion annually to GDP, according to recent research by the Canadian Chamber of Commerce.

I left with a clear conviction. Canada’s approach to AI policy, labour and intellectual property is not a constraint on innovation. It has the potential to become a competitive advantage, but only if it is used to address the industry’s most pressing challenge, the erosion of the creative talent pipeline.

Guardrails as a competitive edge

In 2026, Canada is carving out a distinct position between two global approaches. In the EU, the AI Act has introduced a structured, risk-based framework that can impose heavy compliance burdens on public-facing and immersive technologies. In the U.S., the absence of a unified federal approach has created a fragmented landscape, where companies must navigate a patchwork of state-level rules and shifting standards.

Canada is attempting a third path. Following the stalled progress of Bill C-27, a renewed national push shaped by industry leaders across business, government, non-profit, technology, creativity and culture forums like Banff is focusing on a more agile, sector-specific approach to AI governance. At the centre of this is a growing consensus around creative IP, built on three principles: authorization, remuneration and transparency. AI systems should not use content without permission, creators should be compensated when their work is used and audiences should know when content is AI-generated.

This framework is designed to make Canada a more trusted environment for brands, rights holders and platforms. For companies working with high-value intellectual property, from sports leagues to cultural institutions, that trust is a prerequisite for participation.

The real risk: eating our ‘seed corn’

Policy alone will not determine the outcome. Beneath the surface of the AI boom is a structural risk for the agency and creative industries, the hollowing out of junior talent.

For decades, agencies have relied on early-career roles to build craft through repetition and execution. As AI systems absorb more of that work, the traditional pathway from junior to senior is under pressure. If that middle layer disappears, so does the next generation of creative leadership.

This is the “seed-corn” problem. It is where Canada’s guardrails could have the greatest impact. If AI is deployed purely for efficiency, the industry risks accelerating that erosion. If it is deployed within a framework that values authorship, compensation and transparency, it creates space to rethink how talent is developed rather than replaced.

From prompts to platforms: what agencies need to do next

The implication is clear. Agencies cannot simply adopt AI tools. They need to rethink how they operate.

One shift is moving from prompts to proprietary systems. Rather than relying on off-the-shelf models, agencies can build and train their own brand engines using historical creative data. This ensures that the output reflects their creative point of view, not the generic tendencies of a platform.

Another shift is the rise of clean-room creative environments. By keeping client data and assets within controlled systems, agencies can ensure that proprietary IP is not used to train external models. This is becoming increasingly important for brands operating in regulated or high-value IP environments.

A third shift is in commercial models. As AI compresses production timelines, billing based on hours becomes less relevant. In its place, agencies can explore licensing models, where the value lies in the systems and intelligence they have built.

Canadian organizations are already implementing structured AI governance frameworks that define acceptable risk thresholds, brand parameters and escalation protocols. In high-trust sectors such as payments, companies such as Interac have adopted voluntary AI codes of conduct. This has allowed them to move more quickly while maintaining consumer confidence. The advantage is not just speed. It is the ability to move quickly without breaking trust.

Where this shows up next

As AI-generated content becomes more common, questions of authenticity and credibility will intensify. One response is a shift toward environments where those elements can be more tightly controlled. Physical and experiential formats allow brands to define the rules of engagement, control how AI is deployed and create experiences that feel both immersive and credible.

The opportunity and the condition

Canada has an opportunity to position itself as a leader in AI-driven creative industries, where innovation and protection reinforce each other. That outcome is not guaranteed. Guardrails alone are not a strategy. They are an enabler. The real test will be whether agencies, brands and institutions use this moment to rebuild the systems that develop talent, protect IP and sustain creative value over time.

Alistair Leyland is the founding partner of strategic growth at Array of Stars and is the co-founder and CEO of NVRSE, which aims to shape the next era of digital and spatial experiences.