theme image

Biography

Dr.

Elie Bursztein is a Distiguished Research Scientist at Google & DeepMind. His research currently focuses on creating novel AI-powered cybersecurity capabilities and ensuring AI remains safe and secure for all.

He is currently leading the development of Sec-Gemini, Google's Cybersecurity AI and is involved in the development of Google Secure AI Framework (SAIF). Previously, he incepted password checkup, Google's large scale privacy-preserving service that warns 4B+ users if their passwords are compromised. He also helped create the first SHA-1 collision.

Elie has authored over 60 academic publications, for which he has received more than ten "Best Paper" awards. He has given dozens of talks at premier industry conferences and received multiple industry awards, including a Black Hat Pwnie award.

Passionate about education, he's given countless lectures at universities, including Stanford, Berkeley, and Tsinghua. Elie's work is also regularly featured in major news outlets, including the Wall Street Journal, CBS, Forbes, Wired, the Huffington Post, and CNN.

He is the founder of the Etteilla Foundation, the leading non-profit organization dedicated to preserving and promoting the rich cultural heritage of playing cards. He is also a tarot history expert.

Born and raised in Paris, he received an engineering degree from EPITA and a Ph.D. from ENS-cachan before completing a postdoc at Stanford University. Elie joined Google in 2011, DeepMind in 2023.

Selected Projects

2013

Strengthening Google accounts protections against hijackers and fake accounts

2014

Redesigning Google CAPTCHA to make it easier for humans, resulting in a 6.7% improvement in the pass rate

2015

Deprecating security questions at Google after completing the first large in-the-wild study on the effectiveness of security questions

2017

Finding the 1st SHA-1 full collision

2018

Conducting the first large-scale study on the illegal online distribution of child sexual abuse material in partnership with NCMEC

2019

Developing Keras tuner which became the default hypertuner for TensorFlow and TFX

2019

Inventing Google's password-checking service, Password Checkup

2020

Developing a deep-learning engine that helps to block malicious documents targeting Gmail users

2022

Creating the first post quantum resilient security keys

2023

Magika, a novel AI powered file type detection tool that relies on the recent advance of deep learning to provide accurate detection

Performing

Magic Tricks

Play

Selected

Recent Press

newsletter signup slide

Get cutting edge research directly in your inbox.

newsletter signup slide

Get cutting edge research directly in your inbox.