Andrew Trask

Andrew Trask is a senior research scientist at Google DeepMind, specializing in artificial intelligence and data privacy. His work focuses on privacy and AI, including federated learning, differential privacy, and secure computation. Before pursuing his doctorate, Trask led product analytics at Digital Reasoning, delivering enterprise AI solutions to hedge funds, investment banks, healthcare networks, and government intelligence clients. This experience provided him with insight into the challenges organizations face when managing data responsibly.

Trask is the founder of OpenMined, an open-source community of more than 18,000 researchers developing privacy-preserving AI tools, and the creator of PySyft. OpenMined aims to create an accessible ecosystem of privacy tools and education. Under Trask's leadership, OpenMined has partnered with technology companies including NVIDIA, Google, Meta, Microsoft, and Twitter. The organization also serves as the technology provider for the Christchurch Call Initiative on Algorithmic Outcomes.

Trask contributes to AI policy as a member of the United Nations Privacy Task Force, working to increase awareness and lower barriers for privacy-preserving analytics in the public sector. His keynote presentations address AI safety and governance, introducing technical tools such as homomorphic encryption, secure multiparty computation, federated learning, and differential privacy. Trask demonstrates how these technologies enable organizations to collaborate on sensitive data without exposing it.

As an educator, Trask is committed to making complex AI concepts accessible. He is the author of the book "Grokking Deep Learning," an instructor in Udacity's Deep Learning Nanodegree program, and the author of the deep learning blog "i am trask." He has taught multiple introductory AI courses with more than 12,000 enrollments each, and his deep learning book has sold over 10,000 copies. His lecture on privacy-preserving AI was featured in the MIT Deep Learning Lecture Series. Trask addresses policymakers, technologists, and general audiences on the challenge of building AI systems that are both powerful and trustworthy.