Monday
Where
Common Ground
BSides Las Vegas
Duration
8 hr 52 min
Presents a mixed block of talks on USB dock firmware, encryption policy, post-quantum PKI, drone airspace and dependency risk.
Summary
This Common Ground stream block gathers several short talks rather than a single presentation. The opening segment treats USB docking stations as small embedded computers and shows how a hub controller's firmware update path can be abused through a browser-based updater, since the controller accepts any well-formed 8051 image. A later segment is a policy talk from a long-time digital rights lawyer, who recounts the court fight that freed strong encryption from export controls and argues that privacy acts as a check on power.
Other segments focus on practical engineering. One covers an application security team's attempt to automate CVE fixes and flaky-test repairs in CI, and why interactive LLM sessions with rich context worked better than autonomous bots. A post-quantum PKI lab walkthrough shows how to build a test certificate hierarchy using ML-DSA and ML-KEM on Windows and Linux. A drone airspace talk explains remote ID coverage, sensor limits, and why jamming or kinetic responses are poor choices. A supply chain segment covers malicious package releases, install scripts, and dependency confusion with private registries.
The main takeaways are defensive: pin dependencies, restrict install scripts, monitor critical packages, keep root certificate authorities offline, and treat recurring pipeline failures as a security signal. Depth varies a lot across the block, from policy discussion for general audiences to firmware and cryptography detail.
Related talks
Cryptocurrency Village·30 min
Hanley Shun · Cong Zhang · Oscar Skjerven
Shows how a crypto exchange built an LLM-based scanner to gate code merges, and how internal testers bypassed it.
Briefings·28 min
Bela Genge · Ioan Padurean · Dan Macovei
Who needs AI when raw statistics can do the job just as well—if not better.
Talks·51 min
Vennard Wright
Explores how camera-based AI is used for public safety and where its privacy, bias and legal risks lie.


