Sep 18 2026

Researchers disclose reaching OpenAI's internal monorepo via a forum bug chain, with Claude writing much of the exploit

SAFETY

Original reporting

libheif plus misconfiguration; under 72 hours from discovery to repository access, patched in 14 hours, $6,500 . A documented case of AI-uplifted offense against a frontier lab.

Named in this event

OpenAIAnthropic

Open calls that touch this

OpenAI's Astra maths results survive independent scrutiny · 80%, resolves December 31st 2026
OpenAI never publishes the raw Astra reasoning traces · 75%, resolves February 28th 2027
Anthropic holds the Sonnet 5 price increase · 70%, resolves November 30th 2026
Anthropic IPOs before OpenAI · 75%, resolves August 6th 2027
OpenAI does not IPO in September · 80%, resolves September 30th 2026
OpenAI signs another $10B+ non-Azure compute deal · 65%, resolves August 6th 2027

Issues that mention OpenAI, Anthropic

Around the same time

Sep 18 2026ADOPTPOLICY

Singapore confirms Waymo’s planned path to commercial rides in 2028

Alta

Sep 17 2026CHIPSINFRA

Huawei announces an Ascend 960 supernode using near-packaged optics

Ahuawei

Sep 17 2026MONEYCHIPS

D-Robotics announces a completed $400 million Series C

Afinance

Sep 17 2026SAFETYOPEN

CAISI ranks GLM-5.3 highest among released open-weight models for cyber capability

Anist