Decision Automation
Handing a call that carries judgement to a system, such as which candidate gets a callback or which claim is flagged as fraud. Getting one wrong has a cost a checked task does not, so automating it transfers accountability and needs to be deliberate.
What it means in practice
The task and decision article says a decision is often buried inside a process approved as a single automation, and found only when a result looks wrong and somebody asks who signed it off. It reports that projects stalling after a promising early test were disproportionately ones where a Recommend step had been treated as Execute once the demo looked good. It also notes that the EU AI Act high-risk categories, employment decisions among them, map closely onto Never delegate.
How we use it
The article’s advice is to place each decision on the delegation ladder at scoping, with a named reviewer, a human in the loop, built in from day one. It says a written AI governance policy records which rung each process sits on and who signed off on putting it there.
FAQ
Questions about Decision Automation
How Accurate is AI at making Decisions?
The task and decision article cites a meta-analysis putting AI decision accuracy at around 65% in ambiguous, judgement-heavy situations, against roughly 85% for experienced humans. On structured predictions such as candidate shortlisting it reports about 75 to 80% for AI against 85 to 90% for people. The article reads that as close enough to support the call, not to replace it.
Can AI Support a Decision without making it?
Yes. On the delegation ladder, Assist means AI drafts and compares options while a person decides, and Recommend means AI ranks or suggests while a person stays accountable. The task and decision article says the return at those levels is a better-informed person making the call faster, not a person removed from the loop.
What it means for a Business
Someone still has to explain the outcome when a decision goes wrong. The task and decision article says hiring, dismissal and anything with legal, safety or rights implications stay with a person whatever the technology can do.
Ready to Put This to Work?
Tell us where your team is with AI and we will tell you honestly what would make the biggest difference.