PhD Intern – Responsible AI

Job Description

Research Scientist Intern (Seed Responsible AI) – 2026 Start (PhD) | ByteDance

The Tone:
This is a PhD internship at ByteDance, a global company with research opportunities across the United States, China, and Singapore, dedicated to inspiring creativity and enriching life through innovative products. The company’s Seed team focuses on pioneering new paths toward artificial general intelligence, with research spanning MLLM, GenMedia, AI for Science, and Robotics. This role is crucial for promoting safe and sustainable AI by researching its core mechanisms, providing insights, and offering practical guidance for the organization’s future plans and emerging technologies.

The TL;DR
• Role: Internship (PhD)
• Location: Global (US, China, Singapore)
• Pay: $60 hourly
• Team: Seed-Responsible AI team
• Mission: Promote safe and sustainable AI by researching its core mechanisms to provide insights and practical guidance.

What You’ll Actually Do
• Technology Development: Develop advanced technologies in AI, particularly AI foundation models and AI agents.
• Research Conduct: Conduct research on AI technologies to enhance their reliability, safety, and trustworthiness.
• Product Contribution: Contribute to the development of company products with regards to AI safety and AI ethics.
• Active Contribution: Actively contribute to ByteDance’s products, research, future plans, and emerging technologies.

The Must-Haves
• Background: Currently pursuing a PhD degree in Computer Science, Mathematics, Engineering, or a related field.
• Experience: Strong research background in at least one of these areas: large language model (LLM) training, LLM interpretability, LLM agents, reinforcement learning, or LLM / AI agent safety.
• Skills: Excellent coding ability, data structures, and fundamental algorithm skills, proficient in C/C++ or Python; excellent problem analysis and solving skills; good communication and collaboration skills.
• Bonus: Experience with influential projects or papers in ML, NLP, or LLM is preferred; winners of competitions such as ACM/ICPC, IOI, or TopCoder are preferred.