TooMuchCoding Newsletter - Issue #19 (opens in new tab)

Hi! This week, we’ve got more cybersecurity chaos than a ransomware convention. Anthropic’s unreleased “Mythos” AI model got breached - because nothing says “frontier AI safety” like unauthorized access to your vulnerability-detection superweapon. Meanwhile, the US government is playing whack-a-mole trying to stop AI model distillation, CISA is handing out zero-day patches like candy, and apparently someone actually thought it was a good idea to leave a weather sensor unguarded at an airport ...