Ethics

Applied ethics, in our sense, is the discipline of translating moral reasoning into decisions that institutions can actually make. It is not the same activity as moral philosophy, and it fails in a characteristic way: producing principles that everyone endorses and nobody implements, adopted precisely because they impose no cost and constrain no decision.

There is now an abundance of such principles. Fairness, transparency, accountability, human oversight, beneficence — signed, published, and largely inert. They are inert not because they are wrong but because they are unfalsifiable at the point of use. An engineer choosing a default, a procurement officer scoring a tender, or a ministry adopting a standard cannot determine from them what to do differently. And where a principle cannot change a decision, its adoption is a communications act rather than a governance one. This is the mechanism behind what is usually called ethics-washing, and it does not require anyone to act in bad faith.

We work on the translation problem: turning contested values into specifications, review criteria, procurement conditions, audit tests and refusal thresholds. The test we apply to our own output is deliberately harsh — could an engineer implement this, could an auditor check it, could a regulator enforce it? A principle that fails all three is a sentiment, however widely agreed. We also hold that the deeper structural test is whether acting well depends on individual courage. Where it does, the infrastructure has failed, however admirable the individuals. Courage is not a system.