The right to human intervention is the right to request that a natural person (not an algorithm) review, supervise, or make a decision when you are subjected to artificial intelligence systems, especially if:

• The AI decision has a significant impact on your rights

• Do you think the AI decision is incorrect, unfair, or discriminatory?

• Situation requires contextual understanding that AI does not have

AI Act Art. 14 requires:

High-risk AI systems MUST be designed with human oversight that allows:

Effective human supervision:

• Human can understand AI capabilities/limitations

• Human remains aware of automation (no overconfidence)

• Human can interpret AI output correctly

• Human can decide when NOT to use/follow AI

• Human can intervene/interrupt system

• Human can override AI decision

Difference with entirely human decisions:

AI Only Decision:

• Algorithm decides autonomously

• No humans involved

• AI output = final decision

AI decision making with human supervision (AI Act requires):

• AI provides recommendation/proposal

• Human reviews AI output

• Human has final decision authority

• Human can differ from AI

AI-assisted human decision making:

• Human uses AI as a support tool

• Human maintains full control

• AI does not make decisions

Human supervision levels:

Human-in-the-loop (human in the process):

• Every AI decision requires human approval

• Human reviews first performance

• Maximum control

Human-on-the-loop (human supervises):

• AI operates autonomously

• Human monitors and can intervene

• Intervention when necessary

Human-in-command (human in charge):

• Human can deactivate/modify system

• Human final responsibility

• Strategic control

The AI Act requires at least one of these levels for high-risk systems.