Data Privacy in AI and Machine Learning

#136 - Privacy Engineering: How to Build for Data Privacy - Nishant Bhajaria

Topic 6 of 10

Published Jun 5, 2023 · Clip 39:26–45:46 (6:20)

Nishant cautions against buzzword-chasing (homomorphic encryption, Generative AI, governance) and advises asking questions and getting facts straight. On AI, he says the concern is less about collection volume and more about careless collection without controls; with proper utility checks, deletion after use, and access control, AI data collection can be done thoughtfully. He notes profiling data is also needed for security (distinguishing DDoS attackers from unfairly penalized users) and that AI is built by humans, requiring continuous learning of models, processes, and people.