Ethical Intelligence and the Red Line: Why Power Tests the Line and AI Forces the Question

Ethical intelligence matters most when power is under pressure.

Across democracies, we’re watching what happens when red lines aren’t clearly defined or quietly shifted. In the United States, institutions designed to uphold the rule of law have become recurring sources of stress. Agencies like U.S. Immigration and Customs Enforcement and the U.S. Department of Justice increasingly sit at the centre of public trust debates, not simply over legality, but over legitimacy. Many actions fall within the bounds of law, yet still raise difficult questions about proportionality, transparency, and acceptable consequences.

That distinction matters. Legality is not the same as ethical acceptability. When institutions default to “we’re allowed to do this,” the red line has already shifted.

This is the slow erosion problem.

In The Red Line Reflection, Elizabeth Ayer notes that ethical boundaries rarely collapse all at once. They’re nudged, justified, and reframed, step by step, until consequences feel inevitable rather than chosen. Moral red lines are usually crossed incrementally, not dramatically: each decision seems small, each choice defensible, and the damage only becomes clear in hindsight.

And in Ethics in a Post-Liberal World, the Carnegie Council for Ethics in International Affairs points to a broader pattern: ethical frameworks weaken when power, performance, or preservation become the primary objectives. Norms don’t disappear—they’re deprioritized.

This is the environment AI is entering.

Why AI makes red lines unavoidable

AI systems don’t just optimize actions—they encode values. What data is used, what outcomes are prioritized, what trade-offs are acceptable, and what harms are tolerated all get baked into systems that operate at scale.

As The Future Society argues in What Are Red Lines for AI and Why Are They Important?, red lines are especially critical for AI because:

• AI can normalize practices that would otherwise feel extreme

• Harm may be diffuse, delayed, or invisible to decision-makers

• Accountability can be obscured behind technical complexity

• Once deployed, systems are difficult to unwind

In other words, if you don’t decide your red lines up front, the system will decide them for you.

Ethical intelligence is not about less data

This is where ethical intelligence comes in.

Ethical intelligence isn’t about rejecting data or innovation. It’s about using data to surface blind spots, challenge assumptions, and reveal what’s happening beyond your immediate field of vision.

Used well, data keeps teams honest. It helps identify when engagement is being mistaken for trust, when optimization drifts into manipulation, and when performance metrics mask real harm. Knowing your audience isn’t enough. You need to know who they listen to, how trust flows through networks, and where your choices may be eroding legitimacy, even when the numbers look good.

The red line question leaders avoid

Every organization has red lines. Most just haven’t named them.

Under pressure, political, financial, reputational — unnamed values collapse. Legality becomes the default boundary. But legality is rarely where trust is actually lost.

The harder questions are:

• What outcomes are unacceptable, even if they improve performance?

• What data uses would we refuse, even if competitors adopt them? Who has the authority to stop a system that “works” but feels wrong?

Watching democratic institutions in the U.S. strain under repeated ethical boundary-crossing is a real-time illustration of what happens when red lines are continuously moved in the name of power or expediency.

Canada is not immune, nor are our organizations.

Recent discussions around Alberta separatism illustrate how quickly trust can fracture when institutions are perceived as unresponsive or illegitimate, even when narratives are amplified, well-funded, or strategically propagated. Separatist sentiment doesn’t arise from ideology alone; it often takes hold when people believe systems no longer reflect shared values or fair process. Whether or not separation is realistic is almost beside the point; the signal matters when trust erodes, and people start looking for exits.

Red lines are leadership, not compliance

Red lines are not a communications strategy. They’re not a compliance checklist. They are a leadership decision.

They require clarity before a crisis, not during it. They require human accountability in systems designed to automate. And they require the humility to say: just because we can doesn’t mean we should.

AI makes this unavoidable but the lesson extends far beyond technology. Because the most important question isn’t what performs, or what’s allowed. It’s this: what won’t you do, even if it works?

That’s where ethical intelligence begins.