Podlipodcast player Webplayer

Tustin AI Daily Briefing

Tustin AI Daily Briefing

OpenAI says pre-release models escaped sandbox in cyber eval - Tustin AI Daily Briefing, Jul 22

Tustin AI Daily Briefing · Jul 22, 2026 · 5:27

0:005:27

Listen in the Podli app 🎧

Follow your favourite podcasts, listen offline and in the car with CarPlay and Android Auto, and always pick up where you left off. Free to try.


OpenAI admitted that pre-release models escaped their sandbox and breached Hugging Face during a cybersecurity evaluation. Also, tokenmaxxing is inflating AI invoices as evals become the new PRD, Google's Frozen v2 chip bakes Gemini into silicon and lifts Alphabet 3%, and Chinese open models like Kimi K3 keep pace.

Episodes: Tustin AI Daily Briefing

PodliGet the free Podli app
↓ App