Under Article 3(1) of the EU Artificial Intelligence Act, an “AI system” is a machine-based system that, for explicit or implicit objectives, infers from its inputs how to generate outputs that can influence physical or virtual environments. The definition allows different levels of autonomy and says a system may adapt after deployment; it does not require every system to be fully autonomous or continuously learning.
The legal definition in Article 3(1)
Regulation (EU) 2024/1689—the EU Artificial Intelligence Act—defines an AI system as:
As an Amazon Associate I earn from qualifying purchases.
“a machine-based system that is designed to operate with varying levels of autonomy and that may exhibit adaptiveness after deployment, and that, for explicit or implicit objectives, infers, from the input it receives, how to generate outputs such as predictions, content, recommendations, or decisions that can influence physical or virtual environments”
Recommended Free Tools
Special offer. See more information about Outbyte and uninstall instructions. Please review EULA and Privacy policy.
This is the operative definition in Article 3(1) of the regulation. Its central idea is inference: the system derives from received inputs how to produce an output. The Act does not define AI simply as software that performs a complicated task, nor does it limit the term to systems that generate content.
#1 Best Overall
How inference differs from ordinary rule-based automation
Recital 12 explains that the definition is intended to distinguish AI systems from simpler traditional software and programming approaches. It says systems based on rules defined solely by natural persons to automatically execute operations should not be treated as AI systems on that basis alone. In contrast, inference is a key characteristic: a system determines from inputs how to produce outputs.
The recital mentions machine-learning approaches and logic- and knowledge-based approaches as techniques that can enable inference. These are examples, not a closed list of qualifying technologies. The recital helps interpret Article 3(1); it does not replace the statutory definition.
Rank #2
What the definition does—and does not—require
| Element | What the wording says | What it does not establish as universal |
|---|---|---|
| Autonomy | The system is designed to operate with varying levels of autonomy. | It need not operate independently or without human involvement. |
| Adaptiveness | The system may exhibit adaptiveness after deployment. | Every system need not change, learn continuously, or update itself after deployment. |
| Objectives | The objectives may be explicit or implicit. | The definition does not require objectives to be stated in a particular way. |
| Outputs | Examples include predictions, content, recommendations, and decisions. | The system does not have to produce a decision specifically, or content specifically. |
| Potential effect | Outputs must be capable of influencing physical or virtual environments. | The text does not say the influence must be physical, immediate, or direct. |
These distinctions follow the wording of Article 3(1); they should not be turned into a separate pass-or-fail scorecard.
How to think about borderline software
A product name, marketing claim, or use of a particular programming technique does not settle its legal classification. For a system whose status is uncertain, examine how it works and what its outputs can do:
- Output generation: Does it infer from received inputs how to generate an output, or does it merely execute operations specified by rules defined solely by people?
- Autonomy: What level of autonomy is it designed to have? The definition allows variation, so limited autonomy does not automatically rule a system in or out.
- Adaptiveness: Does it exhibit adaptiveness after deployment? The word “may” means this feature is not stated as mandatory for every AI system.
- Objectives and outputs: What explicit or implicit objectives does it serve, and does it generate outputs such as predictions, content, recommendations, or decisions?
- Potential influence: Could those outputs influence a physical or virtual environment?
These questions organize the statutory language; they are not an official classification test. A specific system may require closer legal and technical analysis.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Check the latest official interpretation
The Act provides for European Commission guidance on applying the Article 3(1) definition in Article 96(1)(f). Because the statutory text does not catalogue every software design or technical example, consult the latest Commission guidance alongside the current consolidated regulation when deciding how the definition applies to a particular system. The EUR-Lex page linked above provides the regulation’s consolidated text; the consolidation identified for this article is dated 27 July 2026.
Quick Recap
Best Value
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
