BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Am I a Member? Auditing Private Machine Learning - Nicolas Paperno
 t\, University of Toronto
DTSTART:20260421T130000Z
DTEND:20260421T140000Z
UID:TALK244585@talks.cam.ac.uk
CONTACT:Alexandre Pauwels
DESCRIPTION:Current privacy evaluations in machine learning (ML) rely pred
 ominantly on membership inference attacks to validate claims of differenti
 al privacy and machine unlearning. By framing ML regulation as a Principal
 -Agent problem\, we demonstrate that regulators cannot rely on such attack
 s alone due to information asymmetry. This can lead to a false sense of pr
 ivacy for individuals whose data is being analyzed. Consequently\, we advo
 cate for a paradigm shift from statistical auditing to algorithmic guarant
 ees. We conclude on the role that cryptography will play for these algorit
 hmic guarantees to be verifiable by third parties.
LOCATION:Webinar &amp\; FW11\, Computer Laboratory\, William Gates Buildin
 g.
END:VEVENT
END:VCALENDAR