WE ARE BUILDING DATA BREACH MACHINES AND NOBODY CARES

Summary from the TLDR Dev newsletter

AI agents are inherently untrustworthy and dangerous due to industry fragmentation and non-determinism, and this might soon show due to the industry's neglect of security during coding with AI.