Meta
Research Scientist Intern, AI Alignment
Menlo Park, CA +4 · Internship · Posted Oct 19, 2026
- Location
- Menlo Park, CA
Seattle, WA
Boston, MA
New York, NY
San Francisco, CA - Experience
- No specific years stated; internship lasts 12 to 24 weeks.
- Education
- Currently has or is in the process of obtaining a Ph.D. degree in Computer Science, Machine Learning, Artificial Intelligence, or relevant technical field.
- Visa sponsorship
- Must obtain work authorization in country of employment at the time of hire and maintain ongoing work authorization during employment.
About the job
Conduct independent, mentored research on AI alignment and machine learning systems, developing and evaluating algorithms and publishing results that may inform Meta products.
What you'll do
- Develop novel algorithms and systems using deep learning, and improve their efficiency, scalability, stability, fairness, security, and privacy.
- Conduct machine learning and AI research, from framing a question through experimentation and analysis.
- Collaborate with research and cross-functional partners, communicate findings, and publish research results.
What they ask for
- Currently pursuing or already holds a Ph.D. in Computer Science, Machine Learning, Artificial Intelligence, or a relevant technical field.
- Experience with a programming language such as Python, C++, C, or Java, and deep learning frameworks such as Pytorch or Tensorflow.
- First-author publication at a listed top-tier ML, NLP, or Systems venue, or equivalent; intent to return to a degree program after the internship.
- Must obtain and maintain work authorization in the country of employment.
Nice to have
- Experience building large-scale machine learning systems and training on large datasets.
- Demonstrated software engineering experience and experience communicating across teams.
Pay
$7,650/month to $12,667/month (base salary)
Tools
Python, C++, C, Java, Pytorch, Tensorflow
AI focus
Research and develop AI alignment and machine learning systems, including work related to reliability, efficiency, evaluation, robustness, safety, security, privacy, reasoning, agents, and alignment with human intent.
Model-written summary · the original posting is authoritative