Where should automated analysis stop?
Routine analytics versus decisions that affect people's working lives
PAWorld27 Zurich · 3–4 March 2027 · Zürich
About this session
One preceding case uses AI to speed up a people analytics team's routine output; the other shows a consequential recommendation that needed a stronger human challenge.
What changes between an automated summary and a recommendation affecting a person? How do we know that review catches material errors rather than approving them?
Learning outcomes
- Classify analytics uses by their consequences for individuals.
- Specify how to test that human review catches material errors.
- Identify who should hold authority to pause or stop an AI-supported process.
Strategic questions
What changes between an automated summary and a recommendation affecting a person? How do we establish that review catches material errors?
Context
Builds directly on 'How a people analytics team shortened the route to a business decision' and 'The AI recommendation that changed after a human challenge'. The same tools can be low-risk in one use and high-stakes in another.
