AI at Overgrad · Safety & Trust

Safety and trust in every AI feature

As Overgrad introduces AI tools, safety and transparency come first. Here's what you can expect from Trek, our AI career guide for students.

In every conversation

  • Built for one purpose

    Trek stays on career exploration and redirects off-topic questions.

  • Unsafe topics are off-limits

    Content rules block discussion of violence, illegal activity, and other unsafe topics.

  • Students make their own choices

    Trek never tells students what they "should" or "must" do.

  • Clear AI disclosure

    Students know they're talking with AI, and AI notes are labeled for educators.

Protecting student data

  • Never used to train AI. Neither Overgrad nor our AI provider trains AI on student data.

  • Always encrypted. Student conversations and files are encrypted when stored and when sent.

  • Only the right people see it. Students see only their own conversations. Educators see only students at their school, based on their role.

  • Minimal personal info. Trek knows a student's first name, grade, interests, and plans, never contact or financial details.

  • Processed in the US. AI runs on Anthropic's Claude models through AWS, in US regions.

  • Deleted on request, following your district's or organization's data agreement.

You stay in control

Districts choose which AI features each school uses. Counselors can review any of their students' full conversations with Trek at any time.

Educator judgment comes first

No system is foolproof. Trek can support student safety, but it doesn't replace your judgment, your school's or organization's protocols, or your relationships with students.

See how Trek works

Questions about AI in your district?

Book time with our team and see how these safeguards work.

Connect with Us