What you’ll actually do
• Own the T&S product surface end-to-end: classifiers, appeals, user-facing warnings, moderator tooling, enforcement flows.
• Own anti-fraud alongside content safety: multi-account farms, free-trial abuse, refund abuse, chargebacks, prompt-injection cost exploits, reseller wrappers. Fraud is a large and growing share of the T&S problem here.
• Translate legal, policy, and financial-risk requirements into concrete product specs the ML and engineering teams can build against.
• Set thresholds and enforcement policy in partnership with Applied ML: what gets blocked, what gets warned, what gets appealed, and the dollar and reputational tradeoffs behind each choice.
• Run the incident loop when something breaks in production, provider escalations, misuse incidents, fraud spikes, new abuse vectors, and ship the fix.
• Design the metrics that matter: dollars saved from fraud, false-positive cost to legit users, false-negative legal exposure, appeal reversal rate, coverage across new models and markets.
• Work directly with Legal, Applied ML, Backend, Growth, Payments, and external providers to keep our contractual and regulatory obligations current.