Author: lie
OpEd on Mythos published in the Globe and Mail
Recently, Bruce Schneier and I wrote an OpEd on Anthropic’s announcement of Claude Mythos Preview, an AI model ...
DetailsJohn’s paper on PSan, a hybrid memory-sanitizer posted!
Ensuring memory safety in unmanaged languages is a significant challenge, as traditional approaches force a difficult trade-off between ...
Details
Highlights on Panel on Digital Transational Repression in the Age of AI
At SRI, we connect technical research to real-world governance challenges. As Director, I get to use my cybersecurity ...
DetailsDiscovering a new vulnerability on confidential computing systems
This past week, we introduced a new attack primitive on confidential computing systems at the USENIX Security 2025 ...
DetailsMy thoughts after the United Nations Workshop on Responsible AI for Peace & Security
Last week, I had the privilege of attending a two-day workshop at the United Nations in New York. ...
DetailsCongratulations to Shawn and John, 2 newly minted PhD students!
A belated congratulations to He (Shawn) Shuang and Shengjie (John) Xu for completing their doctoral dissertations! Shawn’s thesis ...
Details2 TossLab papers at NDSS 2025 next week!
Xiangyu will be presenting his work on EvoCrawl: Exploring Web Application Code and State using Evolutionary Search, which ...
DetailsWelcome Ilya Grischenko to TossLab!
A big welcome to Ilya Grischenko, who is joining us as a postdoctoral fellow from UCSB. Ilya specializes ...
DetailsCongrats to Kexin and Jiaqi for completing their MASc degrees, and welcome to Guozhen and Xiao!
We’ve had a couple of exciting changes this summer going into September. First, both Kexin Li and Jiaqi ...
DetailsExcited to be starting a $5.6M project on Secure and Safe AI
$5.6M Project on AI Safety Funded!
Details