Job Description

About the team

We are part of Global Monetization Product and Technology-Business Integrity-Governance. The Actor Strategy team focuses on securing the ad platform by tackling risks in the advertising ecosystem—such as fraud, impersonation, account takeover (ATO), over-spending, small bad debts, and spam registrations. We balance risk management with user experience to support TikTok Ads' healthy growth. The team continuously enhances its capabilities through platformization, intelligence, and automation, driving smart solutions powered by AI agents, machine learning models, and strategic rules.

Responsibilities:

1. Metric Ownership & Outcome Delivery

- Own core risk metrics: define monitoring definitions, break down targets, conduct trend insights and root-cause analysis, and produce clear conclusions with prioritized action plans.

- Build tiered governance strategies for high-risk industries/regions/attack patterns (e.g., access control, risk controls, enforcement, appeals) and push them into product/system implementation.

2. Strategy Ops & Cross-functional Closure

- Work closely with Product, Engineering, and Algorithms to convert operational risk signals into executable solutions (strategy rules, model requirements, enforcement workflows, platform capabilities), and drive launch + iteration.

- Make strong trade-offs in complex scenarios: strike fraud hard while protecting legitimate advertisers, reducing false positives and ineffective appeals.

3. Evidence-based Governance & Automation

- Promote evidence-based governance: define evidence chain standards and traceable review/enforcement mechanisms to improve interpretability and auditability.

- Leverage traditional predictive models plus LLMs/Agents to explore automated evidence collection, automated investigation, and automated enforcement, upgrading from labor-intensive ops to tech-driven governance.

4. Mechanisms & Incident Response

- Build incident response mechanisms for high-risk events: quickly identify issues, mobilize cross-functional mitigation, run postmortems, and turn lessons into long-term system improvements—raising overall stability and response efficiency.