EU AI Act for Automated grading AI in Education & EdTech
AI used to evaluate learning outcomes maps to Annex III §3(a) — automated grading is high-risk when it determines or materially influences student assessment.
Risk level
Automated grading AI maps to a high-risk Annex III category, so the obligations below apply in full.
Annex III anchor
Annex III, §3
Score basis
A preliminary 74/100 based on the type of decision the system influences and how it is deployed in Education & EdTech.
Provider obligations
What the provider (developer) must do
Deployer obligations
What you must do as the deployer
Deployment
How Automated grading AI shows up in Education & EdTech
Typical contexts
Signals it's in play
- Auto scoring
- Learning outcome evaluation
- Essay assessment
Recommendations
- Teacher review of low-confidence grades
- Bias audits by demographic
- Student appeal process
Watch-outs
- Style over substance
- Demographic grade bias
- Limited-language bias
FAQ
EU AI Act questions about Automated grading AI
Is Automated grading AI high-risk under the EU AI Act?
Automated grading AI maps to Annex III, §3, which the EU AI Act treats as high-risk. In practice it is assessed as High risk, and the obligations below apply to providers and deployers.
Which EU AI Act articles apply to Automated grading AI?
The obligations that typically apply are Art. 9 — risk management including grading bias; Art. 10 — data governance for grading corpora; Art. 14 — teacher oversight of grades before publication; Art. 26 — inform students of AI grading; offer review paths. Providers (developers) carry the technical duties; deployers (operators) carry the use, oversight, and transparency duties.
Who is responsible — the provider or the deployer of Automated grading AI?
Both. Providers owe the technical obligations such as Art. 9, Art. 10. Deployers owe Art. 14, Art. 26. The split matters for procurement and vendor agreements in Education & EdTech.
What should you watch out for with Automated grading AI?
Common failure modes include: Style over substance; Demographic grade bias; Limited-language bias. Mitigations typically start with Teacher review of low-confidence grades and Bias audits by demographic.
Where does Automated grading AI typically appear in Education & EdTech?
Typical deployment contexts include Essay and assignment scoring and Standardised test marking. Before deploying, confirm whether the specific use triggers the high-risk obligations listed above.
Sources
Citations & further reading
Related
More AI use cases in Education & EdTech
AI exam proctoring
Detects and flags suspected cheating behaviour during remote exams.
Read the guideAI admissions screening
Scores and shortlists applicants for admission decisions.
Read the guideEmotion recognition in education
Detects or infers student engagement, attention, or emotions.
Read the guideAdaptive learning platform
Adjusts learning paths and difficulty based on student performance.
Read the guideAI tutoring chatbot
Conversational AI providing subject tutoring and homework help.
Read the guideAI plagiarism detection
Flags copied or AI-generated text in student submissions.
Read the guideLearning analytics dashboard
Aggregates student engagement and performance data for staff dashboards.
Read the guideAI student support chatbot
Handles enrolment, schedule, and service questions for students.
Read the guideAI course recommendation
Matches students to courses based on interests and profiles.
Read the guideExplore
More industry guides
Describe your exact system, get a personalised read
The guide above is a general baseline for Automated grading AI. The free Risk Scanner maps your specific implementation and surfaces hidden compliance blind spots.
Open the Risk ScannerPreliminary EU AI Act clarity summary. Not legal advice.