At Education Perfect, the safety, security, and privacy of our users are incredibly important to us. We’re committed to providing a learning environment that’s just as safe and secure as the classroom.
To help you better understand how we approach these values, especially in relation to our AI-powered features- we’ve put together answers to some of the most commonly asked questions from our community.
Where does AI fit into the EP platform, and what does it do?
For Students: AI Powered feedback Tool
EP’s AI-Powered Feedback tool provides students with instant AI generated feedback on supported extended response questions, which are marked with the AI symbol. Teachers can choose whether to activate this feature or not each time they assign lessons to their students. Teachers have full visibility of both student responses and the AI feedback provided, ensuring they remain central to the learning process. Importantly, this tool is designed to support more effective student learning as part of EP lessons and is not used for assessment.
For Teachers: EP Create
AI is integrated into EP Create to help teachers generate curriculum aligned question sets quickly and easily within EP. If teachers choose to use the AI feature, they choose a curriculum code to write for, and the AI will generate question sets that respond to that curriculum outcome . Teachers can also include additional instructions or guidance for the AI, and can then review, edit and filter the questions before assigning them to their students. Importantly, teachers have full control over the content—nothing is student facing until it's triaged and assigned by the teacher. This ensures teachers remain central to the learning process and that all content is aligned with what they’re teaching in class.
Is EP’s AI-Powered Feedback Tool safe to use in schools?
Yes. EP’s AI-powered features are purpose-built for education. They’re built to be curriculum aligned, with built-in safeguards to ensure it won’t engage with topics that are inappropriate for a school setting. We’ve purposely built guided educational experiences with this technology rather than exposing an open-ended “chatbot”, and teachers have full visibility on all AI generated content.
What is the established process for reporting and correcting AI generated errors or instances of bias?
While EP’s AI models are designed to deliver accurate and unbiased content, no AI is perfect—and occasional errors or instances of bias can occur. Because the system is intended to be teacher-led, we ensure all AI-generated content can be reviewed for quality and accuracy.
To ensure issues are quickly identified and corrected, we have established multiple reporting and correction processes:
Student feedback tools – Students have a built-in feedback button in their learning experience, where they can label responses as incorrect, inappropriate, or unhelpful.
Teacher controls – Teachers can override any AI-generated content, maintaining full control over the learning process.
Direct reporting – Anyone can report concerns through our monitored email inbox at [email protected].
We continuously monitor how teachers interact with generated content—whether it’s kept, edited, or deleted—and use this data to improve the model over time.
What models power your AI feedback?
We use a combination of leading AI models to deliver high-quality feedback. This includes models from Anthropic, OpenAI, Google and AWS, alongside proprietary models we’ve trained ourselves. All models are hosted in our partner data centers located in Canada, ensuring student data remains fully private and secure during inference. Our proprietary models are trained using EP's curriculum aligned lesson material. We DO NOT use any student or teacher data to train our models, and no student or teacher data is ever shared with our model providers.
We regularly reassess which models provide the most accurate marking and the best student learning experience. We've found that different subjects and different tasks work best with different combinations of models. As new models become available, we continually re-evaluate and update our systems to maintain the highest possible performance.
What happens if AI gets it wrong?
While EP’s AI-powered features engage a range of high-quality models designed to deliver accurate and unbiased content, no AI system is perfect—and occasional errors can occur. To minimize this risk, we carefully review and test a range of models to pick the most appropriate one for each task, and we ground the model with curriculum aligned material and instructions. Both students and teachers can flag AI generated content that they disagree with.
We encourage teachers to report any concerns via [email protected] so we can continue to improve the system.
What data will be collected, and how will it be managed?
All data is handled in line with Education Perfect's Privacy Policy. Key points include:
Student Data is stored securely in data centers located in Canada and is never sent outside of Canada. Your data is never sold or distributed.
As an extra layer of precaution, we strongly advise that teachers and students avoid entering personally identifiable information (PII) into AI prompts and we remind them of this when they engage with AI-enabled features.
To support logging and troubleshooting for EP Create, we store limited context related to the AI-generated content—such as the curriculum area, subject, and teacher’s prompt—but this is not retained long term.
Is there an age restriction for students to use the AI features?
No, however as stated above we carefully review and test a range of models to pick the most appropriate one for each task, and we ground the model with curriculum aligned material and instructions to minimize risk.
Does the platform comply with Ontario’s MFIPPA?
Yes. Our platform complies with Ontario’s Municipal Freedom of Information and Protection of Privacy Act (MFIPPA). We ensure that personal information is collected, stored, and used responsibly, in line with MFIPPA requirements for privacy protection and secure data management.
How do you ensure the AI supports equity and inclusion, especially for multilingual or neurodiverse learners?
Our commitment to equity and inclusion is built on a foundation of technical accessibility. Our entire platform is compliant with Web Content Accessibility Guidelines (WCAG) 2.0, as required by the Accessibility for Ontarians with Disabilities Act (AODA). This ensures that students with visual, auditory, physical, and cognitive disabilities can effectively access and interact with our AI tools.
We welcome feedback from both teachers and students—if you have any concerns or suggestions, please don’t hesitate to reach out via [email protected]. Your input helps us continue to improve and ensure a safe, effective experience for everyone.