Is My AI System High-Risk? A Decision Tree
Interactive EU AI Act risk classification decision tree. Covers Article 5 prohibited practices, all 8 Annex III high-risk categories, GPAI model obligations, and Article 50 transparency.
The EU AI Act (Regulation (EU) 2024/1689) sorts AI systems by their intended use: banned practices (Article 5), high-risk systems (Article 6 with Annexes I and III) and systems with transparency duties (Article 50). General-purpose AI models have their own rules (Chapter V). The duties range from outright bans to voluntary codes of conduct.
Use this interactive decision tree to find which rules apply to your AI system. It follows Articles 5, 6, 50 and 51 and Annexes I and III. Answer each question for the intended purpose of your system: the use you state in the instructions for use, promotional or sales materials and technical documentation (Article 3(12)). The result is a starting point, not legal advice.
Interactive Decision Tree
Does your AI system perform a practice banned by Article 5?
Article 5 bans, among other practices: harmful subliminal or manipulative techniques; exploiting vulnerabilities due to age, disability or a social or economic situation; social scoring; assessing the risk of a crime based solely on profiling; untargeted scraping of facial images for facial recognition databases; recognising emotions at work or in education (except for medical or safety reasons); biometric categorisation to infer race, political opinions, trade union membership, religious or philosophical beliefs, sex life or sexual orientation; and real-time remote biometric identification in public spaces for law enforcement, with narrow exceptions. From 2 December 2026, it also bans AI systems that generate sexually explicit deepfakes of identifiable people without their consent or that generate child sexual abuse material (Article 5(1), points (ba) and (bb)).
Understanding the Risk Classification Framework
The EU AI Act takes a risk-based approach: the higher the risk an AI system poses to health, safety and fundamental rights, the stricter the requirements. It applies to providers, deployers, importers, distributors, product manufacturers, authorised representatives and affected persons in the EU, and to providers and deployers in third countries where the system’s output is used in the EU (Article 2(1)). It does not apply to AI systems used exclusively for military, defence or national security purposes (Article 2(3)), to AI systems and models developed and put into service solely for scientific research (Article 2(6)), to research, testing and development before a system is placed on the market or put into service, except testing in real-world conditions (Article 2(8)), or to natural persons using AI in a purely personal non-professional activity (Article 2(10)). AI systems released under free and open-source licences are also outside the Act, unless they are placed on the market or put into service as high-risk systems or fall under Article 5 or 50 (Article 2(12)). The labels “limited risk” and “minimal risk” are common shorthand; the Act itself does not use them.
Prohibited practices (Article 5)
AI practices that the Act bans outright (Article 5), such as social scoring, harmful subliminal or manipulative techniques, exploiting vulnerabilities and certain biometric practices. Fines reach EUR 35 million or, for an undertaking, 7% of total worldwide annual turnover, whichever is higher (Article 99(3)).
High-Risk
AI systems that are a safety component of a product under an Annex I law and must undergo a third-party conformity assessment, or that are used for the Annex III use cases, such as biometrics, critical infrastructure, education, employment, law enforcement and migration. They must meet requirements including risk management, data governance, technical documentation, record-keeping, transparency, human oversight, accuracy and robustness. The requirements apply to Annex III systems from 2 December 2027 and to Annex I Section A systems from 2 August 2028 (Article 113, as amended by Regulation (EU) 2026/1744). Systems placed on the market or put into service before those dates are covered only after significant design changes; in any case, providers and deployers of high-risk systems intended to be used by public authorities must take the necessary steps to comply by 2 August 2030 (Article 111(2)). For Section B systems, only Articles 6(1), 60a and 102 to 112 apply, and Articles 57 to 59 only so far as the sector law has integrated the high-risk requirements (Article 2(2)).
Transparency duties (Article 50)
AI systems that interact directly with people, generate synthetic audio, image, video or text, or perform emotion recognition or biometric categorisation. Providers and deployers must inform people or mark content as Article 50 requires, from 2 August 2026. The information must be clear and given at the latest at the first interaction or exposure (Article 50(5)).
All other AI systems
AI systems for which no Article 5 ban, high-risk rule or Article 50 duty applies. General rules still apply, such as AI literacy (Article 4). Voluntary codes of conduct are encouraged (Article 95).
Common Classification Mistakes
Classifying by label instead of intended purpose
Classification follows the intended purpose: the use the provider states in the instructions for use, promotional or sales materials and technical documentation (Article 3(12)). A system for recruitment, for example to filter job applications or evaluate candidates, is an Annex III use case (point 4(a)). Calling it a suggestion tool does not change that. Only the Article 6(3) conditions can, and never where the system profiles natural persons.
Ignoring the Article 6(3) exception
Not every Annex III system is high-risk. A system that poses no significant risk of harm and meets one of the four Article 6(3) conditions is not considered high-risk. The exception never applies where the system profiles natural persons. The provider must document the assessment before the system is placed on the market or put into service and must still register the system in the EU database (Article 6(4), Article 49(2)).
Forgetting downstream classification
A GPAI model is not itself classified as high-risk. If you build an AI system on a model and that system is high-risk under Article 6, the provider of that system carries the high-risk duties, and the model provider’s duties under Article 53 stay separate. Classify the system by its intended purpose, not by the model alone.
People Also Ask
How do I know if my AI system is high-risk?
An AI system is high-risk if it meets one of two routes. Route 1 (Article 6(1)): it is a safety component of a product, or itself a product, covered by an Annex I law, and that product must undergo a third-party conformity assessment. Route 2 (Article 6(2)): it is used for an Annex III use case and the Article 6(3) exception does not apply; profiling of natural persons is never exempt. For Annex I Section B products, only Articles 6(1), 60a and 102 to 112 apply, and Articles 57 to 59 only so far as the sector law has integrated the high-risk requirements (Article 2(2)). Use the decision tree above to walk through the criteria.
What happens if I misclassify my AI system?
If an Annex III system is misclassified as not high-risk to avoid the rules, the provider is subject to fines under Article 99 (Article 80(7)). Fines for breaking the provider duties in Article 16 reach EUR 15 million or, for an undertaking, 3% of total worldwide annual turnover, whichever is higher (Article 99(4)). For SMEs and small mid-cap enterprises, the lower of the two amounts is the limit (Article 99(6) and (6a)).
Can an AI system's risk classification change over time?
Yes. The Commission can amend Annex III and the Article 6(3) conditions by delegated acts (Article 7, Article 6(6) and (7)). A change of intended purpose can also change the classification: whoever modifies the intended purpose of a system that was not high-risk and is already on the market or in service, so that it becomes high-risk becomes its provider (Article 25(1)(c)). Reassess whenever the purpose or context of use changes. Regulation (EU) 2026/1744 (the “Digital Omnibus on AI”) set the high-risk dates at 2 December 2027 for Annex III systems and 2 August 2028 for Annex I systems. The Article 5 bans have applied since 2 February 2025 (Article 5(1), points (ba) and (bb): from 2 December 2026) and the GPAI duties since 2 August 2025.
Does using a third-party AI API make my system high-risk?
It depends on the intended purpose of your system, not on where the model comes from. If you integrate a third-party AI API into a system with an Annex III use case, such as credit scoring or recruitment, your system is high-risk unless the Article 6(3) exception applies. Whoever places that system on the market or puts it into service under its own name or trademark is its provider (Article 3(3)) and carries the provider duties, whoever built the model. The model provider has its own GPAI duties.
Scanara prepares evidence from your repositories
Scanara reads your connected GitHub repositories and runs rule-based scans for standard AI and machine-learning patterns. It can miss non-standard code. Separately, it generates draft documentation from templates and your assessment answers as evidence for your compliance work. You make the classification decisions.