An AI tutor for Ontario students: what to look for, course by course and board by board

Because in Ontario the code is the course. MCR3U and MCF3M are both Grade 11 functions courses and they are not the same course; MAP4C and MHF4U are both Grade 12 mathematics and prepare for different destinations. A tutor that knows Grade 11 math in general will teach the wrong course confidently. Look for a tool that names the code and sets out the strands, paraphrased from the Ministry description, so you can check that what it thinks the course is matches what your teacher is teaching.

CANtutor AI has a page per code for the main academic sequence: the Grade 9 and 10 mathematics and science courses, the Grade 11 functions, biology, chemistry, physics, English and computer science courses, and the Grade 12 university- and college-preparation courses in the same subjects. Each page links to the wider subject page, where the sticking points are set out in more depth.

The course code tells the tutor what the course is; your teacher's notes, tests and rubrics tell it what your course is. Look for a tool that takes them. In a CANtutor AI session, hold the returned unit test up to the camera and the tutor reads the marked errors itself. Upload the culminating-task brief and its rubric to an assignment workspace, and the per-assignment chat and the AI Grader work from them. A generic tutor answers the topic; a course-specific one answers the document you will be marked on.

Then check the default behaviour. Ask for the answer with no settings changed and see what happens. CANtutor AI starts in Learn mode, which teaches the method and asks you to do the next step; Practice mode waits for your attempt; Answer mode is labelled and warns you at the point of use that turning in AI-written work as your own may breach your school's academic-integrity policy. The grader ignores the mode.

It should be published, it should name the companies that process your content, it should say whether that content trains models, and it should be honest about where it is processed. CANtutor AI's policy is written under PIPEDA and applicable Ontario law, names its providers by company, states that its agreements do not permit them to use content to train general-purpose models, and says plainly that some processing happens outside Canada, including in the United States. Voice-session audio and video are not recorded; only the transcript is kept. A board that requires Canadian-only processing should read that as a no. The full checklist, tied to policy sections, is in the schools article.

Google Classroom connects for personal Google accounts and for school accounts where the district permits third-party apps. Some districts block it; if yours does, everything else still works and you upload your material instead.

In Ontario this is common: many boards' Workspace for Education policies block unapproved third-party apps by default, and the symptom is an access_not_configured error at Google's consent screen. Look for a tool where that is a convenience lost rather than a product broken. In CANtutor AI everything works from uploaded material, and there is a page written to be forwarded to a board's Workspace administrator explaining what the app requests and why. For boards, the Ontario schools page answers the privacy questions boards ask with a link to the policy section behind each answer.

Sessions and generated material spend credits, visible in a per-feature breakdown; the plans and trial terms are on /plans. Start from your course's page at /tutoring/ontario, bring one specific problem from this week's homework to a session, and generate a practice quiz on that narrow topic afterwards. The exam-prep article sets out a plan for the 4U year.