riana

For the first time, starting this year the IEEE Cybersecurity Award for Practice will be given out at Oakland! Nominations are open now for the award, which is intended to recognize individuals and small teams who are generating transformative cybersecurity capabilities that have been realized. Deadline for nominations is February 24. More info here: https://sp2026.ieee-security.org/nominations-cybersecurity-award.html

Cool paid opportunity to participate in important research on ethical best practices for nude image datasets, helmed by Elissa Redmiles, as a follow-up to the same research team's recent NeurIPS paper (https://openreview.net/pdf?id=Ev5xwr3vWh):

I'm on today's episode of The Verge's Decoder podcast with Nilay Patel, talking about, of course, Grok: https://www.theverge.com/podcast/865275/grok-deepfake-undressing-elon-musk-content-moderation

@aram@aoir.social We briefly had one, but then that whole "walled garden duopoly" thing kicked in. https://www.iceblock.app/

This Wednesday at 12pm PT! Come to the book talk I'm hosting with Adam Becker and @rezendi@wandering.shop at Stanford HAI - in-person and Zoom options available! https://hai.stanford.edu/events/adam-becker-and-jon-evans-book-talk

I'm on PBS NewsHour, talking about Grok: https://www.pbs.org/newshour/show/musks-grok-ai-faces-more-scrutiny-after-generating-sexual-deepfake-images

I have an op-ed in the NYT this week about the Grok scandal, sharing my research from last year finding that legal risk hinders AI companies from making their models safer against CSAM - an echo of the years where white-hat hackers were chilled from good-faith research. https://www.nytimes.com/2026/01/12/opinion/grok-digital-undressing.html

@cwg1231@defcon.social Telegram changed its practices real damn fast after the French arrested Pavel Durov

@cwg1231@defcon.social The op-ed mentions a bill that would limit liability for AI model developers as a reward for dataset filtering. There are underexplored Fourth Amendment questions around *forcing* companies to filter datasets (either directly, or by punishing them if they don't), which (I suspect) is why the bill is structured as a carrot, not a stick. And anyway, if Meta can absorb a $5 BILLION fine, anything short of imprisoning execs will get absorbed as cost of doing business, if you're big enough.

Reminder that I maintain a periodically-updated list of papers at the intersection of end-to-end encryption and Trust & Safety. Suggest any I've missed: https://docs.google.com/spreadsheets/d/1aYaDogBF6JegBILgbf--leewyFf-kzkdQwJCjvHTFqI/edit?gid=0#gid=0

ONE WEEK FROM TODAY! Join me at
Stanford HAI for a book talk with Adam Becker and @rezendi@wandering.shop about their respective books, More Everything Forever and Exadelic - both about AI, and both terrific reads, but for very different reasons. Virtual option available! https://hai.stanford.edu/events/adam-becker-and-jon-evans-book-talk

@DaveMWilburn@infosec.exchange Yeah I have no idea whether xAI is doing anything at all to mitigate the NCII or CSAM problems, or whether they would act any differently if given a legal safe harbor; but they're only one company. If we want generative AI to be safer generally, we have to deal with the risks facing the entire field.

I have an op-ed in the NYT today about the Grok scandal, sharing my research from last year finding that legal risk hinders AI companies from making their models safer against CSAM - an echo of the years where white-hat hackers were chilled from good-faith research. https://www.nytimes.com/2026/01/12/opinion/grok-digital-undressing.html

I'm incredibly excited to be hosting two amazing authors & AI thinkers,
@rezendi@wandering.shop and @adambecker@mastodon.social, for a book talk & fireside chat at Stanford HAI on January 21. Please join us in person or virtually! https://hai.stanford.edu/events/adam-becker-and-jon-evans-book-talk

The recording is now online for my recent HAI seminar talk about my research on student-on-student incidents of deepfake NCII: https://hai.stanford.edu/events/riana-pfefferkorn-student-misuse-of-ai-powered-undress-apps?section=watch-event-recording

Cool Job Alert: Clinical Teaching Fellow in the Communications & Technology Law Clinic at Georgetown Law.

Highlights:
* 2-year fellowship for lawyers (JD degree required)
* Start date in July or early August 2026
* Excellent preparation for a career in academia or public interest
* Rolling applications; for priority consideration, apply by December 5
* Supervise JD students on real issues related to law, policy, technology, and justice

https://apply.interfolio.com/177626