AI policy, strategy, ethics and risk

What are the greatest potential benefits and the greatest risks from the development and deployment of artificial intelligence, for individuals, for groups, and for society?

Face Recognition

Creating technical and policy tools:

Co-created face recognition advocacy and policy tools and adversarial attack prototype EqualAIs and Cloak&Pixel, MIT and Harvard

Advising government:

Testified on biometric and facial recognition bills in Massachusetts legislature

Press:

Interviewed by the BBC, WIRED, Science & Vie (Paris), ADA (Berlin)

Why some cities are banning facial recognition technology (video). WIRED

Reconnaissance faciale le grand malaise. Science & Vie (Paris)

Some cities moving to real-time facial surveillance. WIRED

Poll finds Americans trust police use of facial recognition. WIRED

Interviewed on face recognition by Voice of America, aired in Pakistan,

Writing:

The broken shield - face recognition, adversarial attacks, and the dangers of offering imperfect tools. ICLR

Speaking:

Insights on emotion recognition, face recognition, and AI governance. BBC Research, London, U.K.

Facial recognition adversarial attacks, policy, and choice. 7th International Conference on Predictive Applications and APIs, Boston, MA

How machine vision fails: Adversarial attacks and other problems. Re-Work AI in industrial automation summit. San Francisco, CA

The broken shield - face recognition, adversarial attacks, and the dangers of offering imperfect tools," Harvard Berkman Klein Center for Internet and Society

How machine vision fails: Adversarial attacks and other problems. Re-Work deep learning for robotics summit. San Francisco, CA

Unintended consequences of AI (panel). 7th International Conference on Predictive Applications and APIs. Boston, MA

The ethics of face recognition. Teens in AI COVID-19 Online Hack. London, U.K.

Face recognition adversarial attacks, levers of control, and the dangers of imperfect defenses. Yale Technology and Ethics Working Group. New Haven, CT

The broken shield - face recognition, adversarial attacks, and the dangers of offering imperfect tools. Cybersecurity policy working group at the Tufts Fletcher School of Law and Diplomacy. Somerville, MA