As AI becomes embedded in how we work, create and learn, the instinct is to establish rules around its use. But policy that is too rigid risks limiting our ability to understand the technology we are trying to govern.
I believe AI Literacy needs to sit at the centre of policy development. Increasingly, it is also a form of professional and creative agency: the ability to understand these systems well enough to question them, make informed choices about their use and participate meaningfully in shaping the policies that govern them.
Literacy isn’t simply knowing how to use AI. It’s understanding how it works, where it can go wrong, what informs its outputs and when we should, or shouldn’t, rely on it.
Policy should help build that understanding.
Guardrails, Not Walls
We need boundaries around AI, particularly when it comes to attribution, intellectual property, transparency and responsible use. But those boundaries should act as guardrails, leaving room to explore.
I’ve seen this in education. Unrestricted access to AI doesn’t necessarily encourage deeper learning. Thoughtful constraints can. They force students to ask why they’re using the technology, what it’s contributing and whether they should trust the result.
Policy can work the same way.
Rather than trying to prescribe every acceptable use, we can establish clear principles and give people room to develop judgment within them. AI is evolving too quickly for a static rulebook. Our policies need to evolve with it. Adaptive policy does not mean weaker policy: firm principles around consent, attribution, privacy, transparency and accountability can coexist with iterative implementation.
A Seat at the Table
That flexibility also depends on who develops the policy.
Artists, designers and practitioners shouldn’t be brought into the conversation once the rules are largely written. They need to be part of the discourse from the beginning.
We also need translators between professional worlds, that is people who can bridge creative, technical, business and government perspectives. An artist may explain an issue through experience; a technologist through the system; and a policymaker through regulation. If we can’t translate between those perspectives, we risk losing important nuance.
The same is true of consultation. Instead of a linear process of consult, draft and implement, I would rather see us engage, test, re-engage and refine. More touchpoints create better opportunities to understand how policy works in practice before it becomes rigid.
This is why I’m encouraged to see WalterFedy become a founding member of the Canadian Association for the Practice of Architecture (CAPA). CAPA was created in part to give architectural practices a stronger, coordinated voice in the policy discussions that affect our profession.
For me, there is a direct connection to the conversation we’re having about AI. If we believe architects and designers should help shape the policies governing the technologies entering our practice, we also need to create and participate in the mechanisms that give us that voice.
Build Agency, Not Just Compliance
The measure of good AI policy shouldn’t simply be whether people follow the rules. It should be whether those rules help people become informed enough to make better decisions themselves.
Give people guardrails, but also room to explore. Bring practitioners into the conversation early. Create more opportunities for dialogue between those using technology and those governing it. Above all, invest in literacy.
We can’t write a rule for every way AI will evolve. But we can build the criticality to navigate what comes next and make sure our profession has a voice in shaping it.
