2026-09-18

How OpenAI Used Its Own LLMs to Design Its Jalapeño Chip

Also: Security researchers used Claude to help them hack into OpenAI • Researchers used Anthropic’s Claude to hack into OpenAI

Valider

Cache-to-Cache: Direct Semantic Communication Between Large Language Models

Also: Cache-to-Cache: Direct Semantic Communication Between LLMs (2025)

Valider

The Implications of Linguistic Illegibility for LLM Security

Valider

OpenAI launches Astra for Law:GPT-6 legal platform escalates industry race

Also: OpenAI Launches Astra for Law, CombiningGPT-6 Astra with Legal Search and Tools • OpenAI Announces Astra for Law WithGPT-6 Astra, Legal Search Index and Specialist Plugins

Valider

Claude Code now reads AGENTS.md if there is no Claude.md

Also: Claude Code now reads AGENTS.md if there is no Claude.md • Anthropic finally adds AGENTS.md support to Claude Code

Valider

JEV - Can This NEW Type of AI Help Us Win on Kalshi?

Valider

Gavin Newsom is pushing for an AI kill switch

Also: Gavin Newsom signs order to speed creation of anAI'kill switch' - San Francisco Chronicle • Gavin Newsom is pushing for an AI kill switch

Valider

OpenAI 'ethically hacked' with help of Anthropic's Claude chatbot

Also: OpenAI piraté par une IA d'Anthropic: une intrusion orchestrée pour la sécurité - 20 minutes

Valider

Microsoft exec called AI scraping ‘the largest theft of labor in human history,’ new unredacted filings reveal

Valider

Border agents can search cellphones without a warrant or reasonable suspicion

Valider

US Military had close call after using AI for hallucinated intelligence report

Valider

A heap overflow and SSO misconfiguration to compromise OpenAI internal repos

Valider