Security Risks of AI
April 27 2023Stanford and Georgetown have a new report on the security risks of AI—particularly adversarial machine learning—based on a workshop they held on the topic.
Jim Dempsey, one of the...
Read moreUsing LLMs to Create Bioweapons
April 18 2023I’m not sure there are good ways to build guardrails to prevent this sort of thing:
There is growing concern regarding the potential misuse of molecular machine learning models...
Read more
LLMs and Phishing
April 10 2023Here’s an experiment being run by undergraduate computer science students everywhere: Ask ChatGPT to generate phishing emails, and test whether these are better at persuading victims to respond or...
Read moreSide-Channel Attack against CRYSTALS-Kyber
February 28 2023CRYSTALS-Kyber is one of the public-key algorithms currently recommended by NIST as part of its post-quantum cryptography standardization process.
Researchers have just published a side-channel attack—using...
Read morePutting Undetectable Backdoors in Machine Learning Models
February 24 2023This is really interesting research from a few months ago:
Abstract: Given the computational cost and technical expertise required to train machine learning models, users may delegate the task...
Read more
Attacking Machine Learning Systems
February 6 2023The field of machine learning (ML) security—and corresponding adversarial ML—is rapidly advancing as researchers develop sophisticated techniques to perturb, disrupt, or steal the ML model or data. It’s a...
Read moreThreats of Machine-Generated Text
January 13 2023With the release of ChatGPT, I’ve read many random articles about this or that threat from the technology. This paper is a good survey of the field:...
Read moreSerious Security: How to improve cryptography, resist supply chain attacks, and handle data breaches
January 5 2023Lessons for us all: improve cryptography, fight cybercrime, own your supply chain... and don't steal my data and then pretend you're sorry.
Read moreThe bad news: the crooks have your SSH private keys. The good news: only users of the "nightly" build were affected.
Read moreMachine learning security is extraordinarily difficult because the attacks are so varied—and it seems that each new one is weirder than the next. Here’s the latest: a training-time attack...
Read more
Recent Comments