How Learnable uses AI to mark written responses
Student notice for Years 7 to 12
For some predefined Learnable questions, Learnable applies a filter designed to remove common contact details and checks the response for potentially harmful content before OpenAI applies a predefined rubric and suggests a mark. Your teacher can check and override the result.
What happens when I submit a written response?
For eligible questions, Learnable first applies a filter designed to remove common contact details and screens the filtered response for potentially harmful content. If submitted for marking, OpenAI receives the question, filtered response, predefined rubric and marking instructions.
What happens if my response is flagged?
If flagged, you can revise or withdraw your response. To continue, you must confirm that your teacher will be notified. A flag can be wrong; it is not a diagnosis or disciplinary decision. Your teacher or school decides what happens next.
Can I chat with the AI?
No. The AI only applies the predefined rubric to your answer. It cannot chat, answer follow-up questions or provide advice or tutoring.
What personal information is sent?
Learnable does not intentionally send your name, email, school, class, Workspace or account identifier. The filter is designed to remove common contact details but may not catch everything. Do not include private or sensitive information unless your teacher asks for it for an authorised task.
Can AI marking be wrong?
Yes. AI may misunderstand an answer or apply a rubric incorrectly. The mark is proposed only; your teacher can review and override it.
Can I give feedback or ask for a review?
You can select thumbs up or thumbs down. This does not change the mark automatically. Ask your teacher to review a result you think is wrong.
Is my response used to train AI?
Learnable does not opt student responses into OpenAI model training. OpenAI states that API Customer Content is not used to improve its services unless the customer agrees.
How long can the AI provider keep the request?
Under standard API settings, OpenAI may keep prompts and responses in abuse-monitoring logs for up to 30 days. Current settings are listed in Learnable’s Subprocessor Register.
What consent is required?
Before AI marking is used by a student under 18, the school or teacher must provide this notice and obtain any required parent or guardian consent. Contact your teacher to withdraw consent.
How do I report a problem?
Tell your teacher if a mark or warning seems wrong, or if you are worried about something you submitted. Privacy questions can be sent to legal@learnable.education.