Remember 10 years ago, when Netflix had a competition for anyone and everyone to use tons of its user data to come up with an algorithm to better predict viewer preferences? In releasing the data, Netflix said it had scrubbed it of PII, and it was therefore anonymous. But Arvind Narayanan, then a Ph.D student at The University of Texas, and a colleague discovered that the reportedly anonymized data could in fact be de-anonymized, and pretty easily, too. The findings thwarted Netflix's plans for a second, similar contest. In this episode of The Privacy Advisor Podcast, Narayanan discusses life as a researcher at Princeton, including his most recent work on online tracking, which found many sites weren't even aware of third-party tracking occurring on their pages, and that "adult sites" are among the most privacy protective.
02 September 2016
The Privacy Advisor Podcast: Arvind Narayanan
![Default Article Featured Image_laptop-newspaper-global-article-090623[95].jpg](https://images.contentstack.io/v3/assets/bltd4dd5b2d705252bc/blt61f52659e86e1227/64ff207a8606a815d1c86182/laptop-newspaper-global-article-090623[95].jpg?width=3840&quality=75&format=pjpg&auto=webp)
Related stories
Anonymization: The unicorn of privacy engineering
The hidden fragility of AI supply chains: Why traditional risk management falls short
SRB pseudonymization case withdrawn from EU General Court
The value of the DPO in navigating Chile's LPDP
Thought for the week: What a good golf lesson can teach about the impact of geopolitical risk on global data, cyber and AI

This content is eligible for Continuing Professional Education credits. Please self-submit according to CPE policy guidelines.
