Question 1

How many elements does the definition of an AI system according to Article 3(1) AI Act contain?

Accepted Answer

The definition of an AI system in Article 3(1) AI Act contains seven elements that together determine whether a system qualifies as an AI system: machine-based, autonomy, adaptivity, objectives, inference, outputs and environmental influence.

Question 2

What is the difference between autonomy and adaptivity in AI systems?

Accepted Answer

Autonomy refers to the system's ability to operate with a degree of independence. Adaptivity refers to self-learning capabilities allowing the system's behavior to change during use. These are different but closely related concepts.

Question 3

Which element is OPTIONAL in the AI system definition?

Accepted Answer

The use of the term 'may' in relation to adaptivity indicates that a system may, but does not necessarily need to possess adaptivity after deployment to constitute an AI system. A 'frozen' trained model can therefore still be an AI system.

Question 4

What is the most essential and indispensable element that distinguishes AI systems from ordinary software?

Accepted Answer

Recital 12 emphasizes that the ability to infer is an important, indispensable condition that distinguishes AI systems from simpler traditional software systems. Inference refers to the ability to derive from inputs how to generate outputs.

Question 5

Which two main categories of AI techniques enable inference according to the guidelines?

Accepted Answer

Recital 12 AI Act clarifies that the techniques enabling inference include machine learning approaches (that learn from data) and logic- and knowledge-based approaches (that infer from encoded knowledge or symbolic representation).

Question 6

Which four categories of outputs can AI systems generate according to the definition?

Accepted Answer

Article 3(1) AI Act specifies four broad categories of outputs: predictions, content (generated material), recommendations (suggestions for actions) and decisions (conclusions or choices). These differ in the level of human involvement.

Question 7

Which type of system does NOT fall under the AI system definition?

Accepted Answer

Systems for basic data processing that only provide descriptive statistics, hypothesis testing and visualization without learning, reasoning or modeling fall outside the scope. A sales dashboard that only shows averages and trends is not an AI system.

Question 8

What is the difference between the 'objectives' and the 'intended purpose' of an AI system?

Accepted Answer

The objectives of an AI system are internal to the system and refer to the goals of the tasks to be performed. The intended purpose is externally oriented and encompasses the context in which the system is designed to be deployed and how it should be used.

Question 9

Why does the Commission apply a lifecycle approach to the AI system definition?

Accepted Answer

The lifecycle approach recognizes that there is a building phase (pre-deployment) and usage phase (post-deployment), and not every element needs to be continuously present in both phases. For example: inference may appear mainly in the building phase, while outputs occur mainly in the usage phase.

Question 10

A CRM system adds an 'AI assistant' feature. How should this be treated in governance documentation?

Accepted Answer

The underlying application may not fall under the definition, while the AI module does. In governance documentation, it is advisable to describe separately which component constitutes the AI system. This prevents unnecessarily bringing a complete application under AI Act obligations.

AI System Definition Quiz

How many elements does the definition of an AI system according to Article 3(1) AI Act contain?