Lightnews — Scholar-powered news

Yoshi Kohno @yoshikohno.bsky.social · Jul 18

A great @technologyreview.com article quoting from @rachelhong.bsky.social and @willie-agnew.bsky.social!

MIT Technology Review @technologyreview.com · Jul 18

Millions of images of passports, credit cards, birth certificates, and other documents containing personally identifiable information are likely included in one of the biggest open-source AI training sets, new research has found.

A major AI training data set contains millions of examples of personal data

Personally identifiable information has been found in DataComp CommonPool, one of the largest open-source data sets used to train image generation models.

www.technologyreview.com

3 7

Reposted by Yoshi Kohno

Ryan Calo @rcalo.bsky.social · Jun 29

I am so damn excited to share my new book, Law and Technology: A Methodical Approach. Full of stories and priced at $40, this book explains why law finds technology so difficult to regulate, and what to do about it. global.oup.com/academic/pro...

15 59 210

Reposted by Yoshi Kohno

Rachel Hong @rachelhong.bsky.social · Jun 30

New paper alert! In a collaboration between computer scientists and legal scholars, we find a significant amount of PII in a common AI training dataset and conduct a legal analysis showing that these issues put web-scale datasets in tension with existing privacy law. [🧵1/N] arxiv.org/abs/2506.17185

A Common Pool of Privacy Problems: Legal and Technical Lessons from a Large-Scale Web-Scraped Machine Learning Dataset

We investigate the contents of web-scraped data for training AI systems, at sizes where human dataset curators and compilers no longer manually annotate every sample. Building off of prior privacy con...

arxiv.org

1 5 9

Reposted by Yoshi Kohno

Electronic Frontier Foundation @eff.org · Jun 13

Imagine a world where:
💻 Your device is truly yours
📱 You speak freely without fear
🔐 Your privacy is protected
That's the future we're fighting for at EFF. Stand with us for our 35th anniversary!

35 Years for Privacy, Free Speech, and a Brighter Future

Through July 10, new monthly or annual Sustaining Donors get an EFF35 Challenge Coin! With your help, EFF is here to stay.

eff.org

54 220

Yoshi Kohno @yoshikohno.bsky.social · Feb 27

This is @kaimingcheng.bsky.social at his great talk on "User Comprehension and Comfort with Eye-Tracking and Hand-Tracking Permissions in Augmented Reality" at USEC earlier this week. Nice work!! 🎉 Check out the paper at www.ndss-symposium.org/wp-content/u....

2 3

Yoshi Kohno @yoshikohno.bsky.social · Feb 21

Congrats to soon-to-be-Dr Rachel McAmis on passing the UW @uwcse.bsky.social General Exam and advancing to candidacy. Her research, with the UW Security & Privacy Research Lab and the @techpolicylab.bsky.social, is focused on satellites and computer security

Rachel presenting a slide titled "High-level Architecture" in which a ground operator communicates with a ground station, which then communicates with a comms component on a satellite; the comms component is connected to a flight computer, as is the ADCS, power, and payload.

1 6

Reposted by Yoshi Kohno

Aylin Kamelia Caliskan @aylincaliskan.bsky.social · Feb 19

UW’s @techpolicylab.bsky.social and I invite applications for a 2-year Postdoctoral Researcher position in "AI Alignment with Ethical Principles" focusing on language technologies, societal impact, and tech policy.

Kindly share!
apply.interfolio.com/162834
Priority review deadline: 3/28/2025

Apply - Interfolio {{$ctrl.$state.data.pageTitle}} - Apply - Interfolio

apply.interfolio.com

11 11

Reposted by Yoshi Kohno

Umar Iqbal @umariqbal.bsky.social · Feb 18

We are excited to announce that IsolateGPT will appear at NDSS 25. To foster follow-up research, we release our code (github.com/llm-platform...). Also as @llamaindex.bsky.social Pack (llamahub.ai/l/llama-pack...). Lead by Yuhao Wu w/ @franziroesner.bsky.social @yoshikohno.bsky.social & Ning Zhang

GitHub - llm-platform-security/SecGPT: An Execution Isolation Architecture for LLM-Based Agentic Systems

An Execution Isolation Architecture for LLM-Based Agentic Systems - llm-platform-security/SecGPT

github.com

1 3