The CEO’s Guide to AI Governance: Why We Can’t Wait for Ottawa

Posted by Deb Day on February 6, 2026February 6, 2026

Ethical AI isn't a software feature you buy, it's a workflow you enforce. Sharing 5 questions creative leaders need to ask their team.

I spent this past week in Ottawa at Prime Time, the annual gathering of Canada’s media production industry. The optimism in the halls was palpable, but inside the session rooms, we tackled some uncomfortable truths.

Apart from Mark Karney’s positive energy with the Heated Rivalry team and improvised speech, the session that has stayed with me most was "Ethical AI: What It Means in Practice."

The room was filled with producers, funders, and creators, all asking variations of the same questions: When will the rules be clear? Can AI be used ethically? How will government, unions, and guilds intervene?

The panel featured entertainment lawyer Stephen Stohn, Tejas Shah (Gennie), and Marie-Julie Desrochers (CDCE) and was moderated by Kelly Wilhelm. They delivered a wake-up call that every CEO needs to hear:

The cavalry isn't coming. We are the cavalry.

Stephen Stohn put it best, quoting Mark Carney: "We must face the world as it is, not as we wish it to be."

The reality is that Canadian regulation is moving too slowly to keep up with the speed of generative AI. The AI tools that we call on every day (ChatGPT, Gemini, Claude, Perplexity, and others) were built by ingesting vast amounts of information from across the internet. That data is already collected, processed, and embedded in these systems. That reality can’t be undone. Waiting for the perfect AI Act before establishing our own internal guardrails ignores this reality and exposes both individuals and organizations to unnecessary risk.

At Innovate By Day, we believe that ethics isn't a software feature you buy. It is a workflow you enforce. If you are a studio head or marketing leader trying to navigate this "Wild West," here is a 5-step governance checklist, based on the hard truths from the session:

Here is what the report signals for our industry partners and what it means for the future of how content gets made, positioned, and discovered.

The Reality: If an AI generates your script, character design, game mechanics, or soundtrack, your legal claim to that work may be far weaker than expected… or may not exist at all. As Stephen Stohn warned, there is currently a "copyright void" for machine-generated content.

The CEO Question: “Can we prove a human created the core value of this asset?”

If you are building a brand or a franchise on AI-generated foundations, you are building a castle on sand. You need a human in the loop, not just for quality, but for ownership.

The Reality: There is a growing temptation to use AI to “improve” historical records, whether that be to smooth a grainy 1970s recording, clarify a blurred image, or enhance an imperfect transcript. But these alterations don’t just clean up the past; they subtly change it. When factual accounts are reinterpreted or altered, even with good intentions, they undermine the integrity of the historical record. Over time, these modified versions risk being treated as primary sources. Future AI systems may then train on these inaccuracies, compounding errors and accelerating the spread of misinformation, further eroding our ability to verify what is real and what is not.

The CEO Question: “Are we prioritizing polish over truth?”

Especially for those of us in documentary and non-fiction, our duty to the truth outweighs our desire for a slicker edit.