Trust & Safety Lead, Education at Playlab · LumiJobs

Responsibilities

Playlab is looking for a Trust & Safety Lead to own safety on our platform, with a focus on children and young adults. We’ve built Playlab with responsibility in mind, but this work is an ongoing commitment to the students and educators we serve.
You'll make the calls on live incidents, moderation escalations and safety policy. Just as important, you'll help design the systems that catch problems automatically, so people step in only where human judgment matters.
Educators use Playlab to build AI apps for their own classrooms. That means our safety work covers two things: how the AI behaves, and what people build and do with it. Responsible AI in education is a young field, so many of the playbooks you'll run haven't been written yet. You'll write them.
This role owns the decision. When something happens, you decide what to do: suspend an account, take down an app, notify a school, file a required report. You'll bring in leadership for the rare calls that need it, and keep the relevant folks informed about what you did and why.
You'll report to our CTO and work closely with product, engineering, learning engineering and partnerships, as well as the people leading our responsible AI and privacy work.
Build the systems that catch problems
Learn what "bad" looks like here. Review flagged content, hunt for misuse our filters miss, and try to build apps that break our policies. Stress-test features before and after they ship, and grow a community of internal and partner testers who do this with you.
Define what our systems should catch. Build a Playlab-specific harm taxonomy and clear labeling guidelines, so people and systems judge content the same way.
Turn what you find into examples we can build on. Collect what you see in reviews, hunts and red teaming into labeled sets engineering can use to build and test detection.
Write the requirements. Spell out what our filters, rules and signals need to catch and what good looks like. Engineering builds it; you decide whether it's working.
Track how well detection works. How much we catch automatically versus what a person finds, how often we flag things that are fine, and how long problems go unnoticed. The goal is for people to spend their time on the calls that need human judgment.
Own incidents end to end

Trust & Safety Lead, Education

Responsibilities

Qualification

Preferred

Benefits

Company

Playlab

More roles at Playlab