Regional AI hyperspecialization
Policy Brief - Reading Group
How smaller nations can stay strategically relevant by specializing in narrow, high-value AI capabilities — by Tomas Felipe Rubio Diaz and Hojun Jin.
The CASI Blog is where we share what's happening across our CASI community - from research findings and reading group discussions to events, speaker talks, and opportunities in AI safety. As CASI engages with students and researchers across CMU and beyond, this blog is how we keep our community informed, connected, and inspired to take the next step in their AI safety journey.
Three federal liability frameworks
CASI Policy Hackathon Finalist 2026
Strict liability, negligence-based duty of care, and an industry-funded compensation model for AI-enabled consumer fraud — by Dorian Clay and Jessica Wen.
AI-powered FRT and civil liberties
CASI Policy Hackathon Finalist 2026
A Fourth Amendment + Rawlsian ethics framework for permissible use of AI-powered facial recognition at the border — by Taeho Lee.
AI voice-cloning fraud
CASI Policy Hackathon Finalist 2026
Know Your Deployer (KYD), conditional cost-shifting, and a hybrid BSA–TILA model for agentic AI platforms — by Mateo Ruiz-Leal.
AICCLA
CASI Policy Hackathon Finalist 2026
A proposed AI Creator Compensation and Licensing Act: blanket licensing, royalty tiers, federal registry, and safe harbor — by Jong Hyun Son and Soomin Seo.
AAAI-26
Policy Brief - Reading Group
Presenting AI safety research on jailbreaking and chain-of-thought patterns at AAAI-26