Safety component
A part of a product that stops people from getting hurt, if it fails, someone could be harmed. When AI acts as this kind of component, it falls under strict safety rules.
The Simple Version
A part of a product that stops people from getting hurt, if it fails, someone could be harmed. When AI acts as this kind of component, it falls under strict safety rules.
Detailed Explanation
Under Article 6(1) of the EU AI Act, an AI system that is a safety component of a product covered by a Union harmonisation law listed in Annex I, such as machinery, medical devices, or automotive vehicles, is classified as a high-risk AI system regardless of whether it appears in Annex III. This provision captures AI embedded in physical products where incorrect AI output (e.g. a misclassification by an AI-powered braking system) could cause physical injury. Safety component AI must comply with both the sector-specific product regulation and the AI Act's high-risk requirements.
Key Characteristics
- Defined in relation to Annex I sector regulations (machinery, medical devices, etc.)
- AI acting as safety component is automatically high-risk under Article 6(1)
- Dual compliance required: sector regulation and EU AI Act high-risk requirements
- Failure can directly cause physical harm to persons
Why It Matters
Hardware and product manufacturers integrating AI into safety-critical components (braking systems, patient monitoring, industrial automation) must understand that the AI component triggers EU AI Act high-risk obligations in addition to sector-specific CE marking requirements.
Real-World Analogy
Like the airbag sensor in a car, a component whose correct functioning is essential for preventing injury. If that sensor is AI-powered, it becomes a safety-critical AI component subject to both automotive safety standards and AI Act requirements.
Common Misconceptions
- Only standalone AI products are high-risk. AI embedded in physical products as safety components is equally high-risk.
- Meeting the sector product regulation is sufficient, safety-component AI must satisfy both the sector regulation and EU AI Act high-risk requirements.