Skip to content
WyteRock E-Books

Free self-check

Is your AI tool high-risk under the EU AI Act?

Eight questions, about three minutes. You get a plain orientation (likely prohibited, likely high-risk, transparency duties or minimal risk) with the articles that apply and the dates they apply from. Nothing you enter leaves your browser unless you ask for the result by email.

Does the tool do any of the practices banned by Article 5?

Harmful manipulation or deception; exploiting age, disability or social or economic situation; social scoring; predicting crime from profiling alone; untargeted scraping of facial images; emotion recognition at work or in education; biometric categorisation inferring race, politics, religion or sexual orientation; real-time remote biometric identification in public spaces for law enforcement.

Is it a safety component of a regulated product, or itself such a product, that needs third-party conformity assessment (Annex I)?

For example a medical device under the MDR or an in vitro diagnostic under the IVDR assessed by a notified body, or machinery, toys, lifts, radio equipment.

Is it used in one of the Annex III areas?

Biometrics; critical infrastructure; education and vocational training; employment and worker management; access to essential services, including credit scoring and risk pricing in health and life insurance; law enforcement; migration, asylum and border control; administration of justice and democratic processes.

Does it profile people, that is, evaluate or predict aspects of a person such as work performance, health, preferences, behaviour or location?

Under Article 6(3), an Annex III system that profiles people is always high-risk.

Does it only do a narrow procedural task, improve the result of work a person has already done, detect patterns without replacing human assessment, or prepare an assessment?

These are the Article 6(3) conditions under which an Annex III system may not be high-risk.

Does it talk to people directly (a chatbot or voice assistant), generate or alter images, audio, video or text (including deepfakes), or recognise emotions or categorise people biometrically?

These trigger the transparency duties of Article 50.

Are you placing a general-purpose AI model on the EU market yourself (not building on someone else's model)?

Models such as large language models that can serve many tasks. Building an application on a third-party model does not make you its provider.

What is your role?