AI Use Policy
Last updated: 2026-01-01 · Version 1.0-draft
⚠ [ATTORNEY DRAFT REQUIRED] — This page contains placeholder legal text that has not been reviewed or approved by a licensed attorney. Do not rely on this content for legal compliance. Replace all placeholder sections before production launch.
1. What AI Is Used For
GradeWise uses artificial intelligence to:
- Generate per-criterion grade suggestions and score ranges for student essays
- Draft written feedback suggestions for teacher review
- Surface instructional recommendations based on student skill patterns
2. What AI Is Not Used For
- No AI makes final grading decisions. Every AI-generated grade is a suggestion that must be reviewed and approved by the teacher before it is recorded.
- No AI takes action without teacher approval. The system enforces human review at every consequential step.
- AI does not communicate with students. AI-generated feedback is delivered to the teacher, not directly to students.
3. Model Providers
GradeWise uses the OpenAI API for essay grading and feedback generation. The specific model version used is configurable. We do not use fine-tuned models trained on student data.
4. Student Data and AI Training
[VERIFY CURRENT OPENAI API TERMS] — Student essay content is sent to the OpenAI API for grading. OpenAI's API terms of service prohibit using API inputs for model training by default. We do not opt in to any training data sharing with OpenAI or any other provider. For current OpenAI data usage terms, see OpenAI API Data Usage Policies.
5. Human Oversight Requirement
Every AI-generated grade must be reviewed by the teacher before it is considered final. GradeWise enforces this at the system level: grades cannot be shared with students or recorded as official until the teacher explicitly locks them after review.
6. Accuracy and Errors
AI grading is assistive, not authoritative. The teacher is responsible for all final grades. If the AI produces an incorrect or inappropriate grade or feedback suggestion, the teacher's override is the correction mechanism. We do not guarantee the accuracy of AI-generated content.
7. Bias and Fairness
AI rubric-based grading is subject to the same biases embedded in the rubric itself. Teachers are responsible for the rubrics they define. We do not provide automated bias detection. Teachers should review grade distributions for unusual patterns and adjust rubrics accordingly.
8. Updates to AI Use
Material changes to our AI providers or use cases will be disclosed to account holders with at least 30 days' notice before taking effect.
Contact
Questions about our AI use? Contact us at [email protected].