Skip to content
April 2026 produced two events that, taken together, capture a hard question in AI ethics: who gets to decide the guardrails, and what happens to people’s rights while that is argued out.
Washington versus a state
xAI challenged Colorado’s AI Act in federal court, and the US Department of Justice intervened on the company’s side, with the state agreeing to suspend enforcement. A federal government actively helping to unwind a state law aimed at algorithmic discrimination is a landmark moment, and a sign of how contested the direction of AI regulation has become.
A court clears facial recognition
In the same month a UK court upheld the police use of live facial recognition, finding the current approach lawful even though it rests on policy rather than a dedicated statute. Campaigners are appealing. The technology is being scaled up while the legal foundation is still being tested.
The ethics underneath the law
Regulation will keep swinging between prescriptive and permissive, and between federal and local. What does not change is the underlying duty to the people affected: a candidate who is screened out, a passer-by whose face is scanned. Waiting for the law to settle is not a strategy, because the harm, if there is one, happens in the meantime.
Our view is simple: build and use AI so that it would stand up whether the rules tighten or loosen. That is what the Idiro AI Ethics Centre is for.
Sources: DOJ intervention in xAI v Weiser · UK police facial recognition ruling (Pinsent Masons)
This article is general information, not legal advice.
