Product Manager, Red Team, Platform Responsibility

Role Overview

This role sits within the Red Team function focused on evaluating content safety and security systems from an attacker’s point of view. The goal is to uncover weak spots, surface new abuse methods, and recommend improvements so the platform remains effective across different languages and regions.

Team Focus

The team tests global content security controls by simulating adversarial behavior, examining how systems respond to abuse, and turning those findings into practical recommendations that strengthen moderation and safety outcomes.

Key Responsibilities

• Identify threat patterns by reviewing signals from TikTok’s internal environment, external escalations, competitive intelligence, and threat intelligence sources. Track emerging abuse methods and evasion tactics, and keep the team’s adversarial knowledge base current through structured frameworks such as evasion taxonomies and abuse pattern models.
• Convert raw indicators into clear, actionable threat pattern records that can be used by the broader team.
• Own the sample-construction workflow end to end, starting from threat intake and continuing through the creation of realistic adversarial samples at scale.
• Set the methods, tooling needs, and quality expectations for sample building, while ensuring coverage aligns with the Abuse × Evasion attack surface.
• Design and run adversarial evaluation programs that measure how well the platform withstands realistic attacks.
• Define evaluation frequency, coverage benchmarks, and reporting processes, then use results to feed improvements into policy, model, and operations teams.
• Lead the development of LLM-based agent systems that help automate Red Team workflows, including signal triage, evasion-pattern matching, abuse-pattern analysis, and sample-construction support.
• Establish agent workflows, human review checkpoints, and automation quality standards, while continuously testing how far AI-assisted adversarial analysis can be scaled.

Qualifications

The ideal candidate should bring a strong background in content security, trust and safety, or adversarial testing, along with experience working in large-scale consumer platforms. A solid understanding of content moderation systems is important, as is practical experience using AI or LLM tools for agent design, prompt engineering, or workflow automation. The role also requires the ability to manage complex programs with multiple moving parts and cross-functional dependencies.

Preferred Experience

Additional value will come from experience building AI agent pipelines for operational automation, especially for triage, classification, and structured output generation. Familiarity with adversarial techniques used against machine learning systems, such as evasion, prompt injection, and multimodal attacks, is also useful. Experience with threat intelligence, OSINT, or dark web monitoring in a trust and safety setting, as well as knowledge of international content security regulations and regional policy compliance, would be advantageous.

About TikTok

TikTok is a global short-form video platform with a mission to inspire creativity and bring joy. Its headquarters are located in Los Angeles and Singapore, and it operates across major cities worldwide.

Why This Role Matters

This position helps strengthen the platform’s defenses by finding weaknesses before attackers can exploit them, improving safety systems, and supporting the teams responsible for policy, machine learning, and operations.

Diversity & Inclusion

TikTok emphasizes an inclusive workplace where people are valued for their abilities, backgrounds, and perspectives, and where diverse voices help shape the experience for communities around the world.