Why AI governance needs the moral vocabulary of the oil patch

With the launch of the federal government’s post-Artificial Intelligence and Data Act (AIDA) regulatory framework and Alberta’s Artificial Intelligence Data Centres Strategy, the conversation has shifted. While these are necessary for the “innovation economy,” they are ethically thin. We are racing to build the infrastructure of the 21st century without the shared moral vocabulary that made the oil era so stable and so culturally potent. AI arrives in a different moral ecosystem. It is anchored not in churches, but in research labs and corporate campuses. Unlike oil, which was viewed as a finite gift to be managed, AI is treated as endlessly expandable, a digital alchemy. But part of what makes AI unsettling is precisely that it has no religious leader(s). There is no pope of machine learning, no defined or appointed council of elders with the authority to excommunicate a dangerous model or demand repentance after harm. Instead, we have billionaire founders who oscillate between apocalyptic warnings and utopian promises, answering mainly to boards and markets. Where oil in Alberta had pastors and premiers ready to claim oil was a gift from God, AI has no equivalent institution capable of saying, “This is wrong, even if it is profitable.” Our current policy approach, such as the federal Directive on Automated Decision-Making, relies on “Algorithmic Impact Assessments.” These are technical audits designed to catch bias, but they cannot provide a moral north star. They can tell us if a model is discriminatory, but they cannot tell us if the very goal of the model is consistent with a “good life” for Canadians. This absence of moral authority is what makes our handling of AI precarious. Religion, for all its faults, at least offers shared stories, named virtues, and a language for collective guilt. The AI ecosystem operates with immense power but almost no shared public moral framework. Even as Canada moves toward digital sovereignty in 2026, we remain a country with a sophisticated tech policy but a hollowed-out moral core. Alberta’s history proves that when an economy fuses with a single moral story, it gains immense momentum, but it also risks blindness. As Canadian regulators, Indigenous nations, and universities look to the next phase of AI governance, they must resist the urge to treat AI as a purely technical challenge. We must move beyond “safety audits” to “value audits.” The question is whether Canadian institutions will insist on bringing human moral traditions into the governance of AI, or whether we will let a new, godless creed of optimization decide our future. The real choice is whether we invite our deepest, plural moral vocabularies into the policy room, or leave it to the markets alone.