AI Glossary
Human Oversight
Human Oversight is the practice of ensuring people remain involved in monitoring, reviewing, or approving important AI decisions.
Human Oversight
Overview
Artificial intelligence can automate many tasks, but some decisions still require human judgment.
This principle is known as Human Oversight.
Human Oversight is the practice of ensuring people remain involved in reviewing, supervising, or approving important AI decisions.
The amount of oversight depends on the level of risk.
Low-risk AI systems may require occasional review.
Higher-risk systems often require continuous human involvement.
A helpful way to think about human oversight is a commercial airline pilot.
Modern aircraft can fly much of the journey automatically.
However, pilots continuously monitor the aircraft and are ready to intervene whenever necessary.
Organizations use AI in a similar way.
AI assists people.
It does not eliminate human responsibility.
Human oversight works alongside AI Accountability, AI Governance, and Responsible AI to help organizations deploy AI safely.
Why It Matters
Keeping people involved helps organizations detect errors, reduce risk, improve trust, and ensure important decisions remain accountable.
Real-World Example
A financial institution uses AI to evaluate loan applications.
Instead of allowing the AI to approve every application automatically, loan officers review high-value or complex applications before making the final decision.
Related Concepts
Related Articles
- AI Governance vs AI Risk Management
- What AI Governance Actually Means
- Responsible AI Without Fearmongering