Multimodal AI PhD Intern (Summer 2026)
Role Overview
This is a PhD-level internship where you'll conduct cutting-edge research on multimodal deepfake detection, focusing on audio and vision modalities. You'll work with the AI team to propose new methods, implement and evaluate ideas using Python and PyTorch, and publish results in peer-reviewed journals. The role involves independent research and collaboration, aiming to advance the company's detection platform and contribute to academic publications.
Perks & Benefits
The internship is fully remote with the option to work from the New York City HQ, offering flexibility in location. It provides hands-on experience with modern deep learning tools like AWS/GCP and opportunities for career growth through publishing research and working with experienced AI researchers. The culture emphasizes innovation and mission-driven work in cybersecurity, with a focus on real-world impact against AI-generated media threats.
Full Job Description
Who we are.
Reality Defender is an award-winning cybersecurity company helping enterprises and governments detect deepfakes and AI-generated media. Utilizing a patented multi-model approach, Reality Defender is robust against the bleeding edge of generative platforms producing video, audio, imagery, and text media. Reality Defender's API-first deepfake detection platform empowers teams and developers alike to identify fraud, disinformation campaigns, and harmful deepfakes in real time.
Backed by world class investors including DCVC, Illuminate Financial, Y Combinator, Booz Allen Hamilton, IBM, Accenture, Rackhouse, and Argon VC, Reality Defender works with leading enterprise clients, financial institutions, and governments in order to ensure AI-generated media is not used for malicious purposes.
Youtube: Reality Defender Wins RSA Most Innovative Startup
The Multimodal AI Internship.
The 4-month internship is designed for current PhD students and candidates to partner with Reality Defender's AI team to conduct cutting-edge research and publish peer-reviewed papers. Your primary collaborators will be Surya Koppisetti and Yi Zhu, who will guide and advise your efforts within multi-modal deepfake detection. This internship can be performed remotely, although you're welcome to work from our HQ in New York City.
What you'll do.
Investigate and propose new methods for detecting generative multi-modal content, spanning audio and vision modalities.
Perform research on multi-modal deepfake detection and reasoning tasks.
Collaborate with researchers in the team.
Write up results of research for internal reports and submission to academic journals/workshops.
Independently implement and evaluate ideas on modern deep learning stack - Python, PyTorch, and GPU-enabled cloud compute, like AWS/GCP.
Who you are.
PhD student in a relevant technical field, preferably three or more years into the program
Experience in multi-modal learning, such as in audio-visual classification and audio-language reasoning.
Proficient in Python and in building deep learning models with PyTorch.
Published peer-reviewed research papers in reputable AI and speech venues, e.g. CVPR, NeurIPS, ACL, Interspeech.
Excited about Reality Defender's mission to build a best-in-class and comprehensive deepfake and AI-generated content detection platform.
Available to start May 1st, 2026, for a minimum duration of 4 months.
Similar jobs
Found 6 similar jobs