Research Scientist Intern, Security & Integrity (PhD)
Description
Reality Labs at Meta is building the next generation of wearable and AR/VR devices, and the platforms behind them. Our Trust team keeps the people who use these products safe. We work on two connected problems. One is the security of our devices and systems, across hardware, firmware, operating systems and software. The other is the integrity of our platforms against abuse, fraud and adversarial actors. We combine research, data and production engineering to find high-risk areas, build new detection and mitigation techniques, and prove they work before scaling them. We're looking for PhD researchers in security, privacy, integrity or related fields who want to protect people and products at scale. That could mean securing devices and systems against attacks, or detecting and stopping abuse and fraud on our platforms. Interns own a well-scoped project from start to finish: defining the problem, building a working prototype or system, measuring results, and sharing what they learned. Our internships are fourteen (14) to twenty-four (24) weeks long, with various start dates throughout the Spring, Summer and Fall.
Responsibilities
Conduct research on security or integrity challenges across Meta's products, platforms and devices. Own a well-scoped research project end to end, from problem definition and baseline measurement through a working prototype, system or launched experiment with measurable results. Design, prototype and evaluate new techniques for detecting and mitigating threats. Examples include vulnerability discovery (fuzzing, program analysis, emulation, formal methods) and detecting malicious or adversarial actors (ML models, verification systems, data pipelines). Define metrics for measuring harm or risk, and evaluate solutions for precision, coverage and impact against existing defenses. Research and apply state-of-the-art techniques, including AI and agent-based tools, to scale security and integrity analysis. Work with cross-functional partners (engineering, product, policy, data science, hardware, infrastructure) to turn research into practical defenses. Communicate findings through design documents, technical presentations and, where appropriate, publications.
Qualifications
Currently pursuing or recently completed a PhD in Computer Science, Electrical Engineering, Cybersecurity or a related field Research experience in security, privacy, integrity, trust and safety, adversarial machine learning, or a related area Programming experience in Python, C/C++, Java or a similar language Demonstrated ability to conduct independent research and deliver technical results Strong analytical and problem-solving skills for tackling new adversarial challenges Excellent communication skills for collaborating across interdisciplinary teams Must obtain work authorization in the country of employment at the time of hire, and maintain ongoing work authorization during employment Intent to return to a degree program after the internship Proven track record of interdisciplinary research collaboration and delivering research prototypes Experience with one or more of the following: Systems or hardware security: low-level and embedded systems, ARM architecture, Linux, trusted execution environments (e.g., ARM TrustZone, Intel SGX, AMD SEV), secure boot, or emulation (e.g., QEMU) Building or extending security testing tools such as fuzzers, static or dynamic analysis, or binary analysis frameworks Abuse, fraud or adversarial-actor detection, including building ML models and dataset pipelines and handling issues such as data leakage Large-scale data analysis with SQL, and designing or analyzing A/B experiments Demonstrated ability to integrate AI tools into research or engineering workflows to drive measurable impact Track record of significant results, as shown by grants, fellowships, patents, or first-authored publications at leading conferences or workshops
Compensation: $7,650/month to $12,667/month + benefits