Waymo Announces Singapore Expansion Targeting 2028 Ride-Hailing Launch1h ago Two Axes of Alignment: A Framework for Robust Superintelligence Alignment1h ago Three Hackers used Opus 5 to Hack Into OpenAI's Core Codebase [WSJ]2h ago Anthropic Says Claude Drives 26% of Its Research and Development2h ago Chinese Robotics Suppliers Rise on Tesla Optimus Audit Report2h ago BREAKING NEWS: BOJ to watch forex moves, AI-linked demand for future monetary policy2h ago Deep recurrent models are less robustly CoT-monitorable than normal CoT models in a toy setting3h ago Will machine intelligence kill human expression?4h ago Waymo Plans to Offer Paid Robotaxi Rides in Singapore in 20284h ago The U.S. and China Want AI Guardrails. But Their Ideas Couldn’t Be More Different.4h ago The Cost of Utopias (a Dialog)4h ago TypeSafe AI exits stealth with $40M to build AI for use by software4h ago Microsoft exec called AI scraping ‘the largest theft of labor in human history,’ new unredacted filings reveal4h ago Indian Project Financier’s Data Center Loans Top $1.2 Billion4h ago

Waymo Announces Singapore Expansion Targeting 2028 Ride-Hailing Launch1h ago Two Axes of Alignment: A Framework for Robust Superintelligence Alignment1h ago Three Hackers used Opus 5 to Hack Into OpenAI's Core Codebase [WSJ]2h ago Anthropic Says Claude Drives 26% of Its Research and Development2h ago Chinese Robotics Suppliers Rise on Tesla Optimus Audit Report2h ago BREAKING NEWS: BOJ to watch forex moves, AI-linked demand for future monetary policy2h ago Deep recurrent models are less robustly CoT-monitorable than normal CoT models in a toy setting3h ago Will machine intelligence kill human expression?4h ago Waymo Plans to Offer Paid Robotaxi Rides in Singapore in 20284h ago The U.S. and China Want AI Guardrails. But Their Ideas Couldn’t Be More Different.4h ago The Cost of Utopias (a Dialog)4h ago TypeSafe AI exits stealth with $40M to build AI for use by software4h ago Microsoft exec called AI scraping ‘the largest theft of labor in human history,’ new unredacted filings reveal4h ago Indian Project Financier’s Data Center Loans Top $1.2 Billion4h ago

Cybersecurity1 min reading time

Three Hackers used Opus 5 to Hack Into OpenAI's Core Codebase [WSJ]

LessWrongPublished 2 hours ago

Three hackers exploited the Opus 5 tool to breach OpenAI's core codebase, according to a report by the Wall Street Journal. The incident raises concerns about security vulnerabilities in AI development environments.

#codebase-hacking #openai #opus-5 #security-breaches #wall-street-journal

More in Cybersecurity

Hackers Used Anthropic’s Claude to Break Into OpenAI

The Wall Street Journal5h ago

Cybersecurity5 min read

The fix for rogue AI agents could be more AI

TechCrunch9h ago

Cybersecurity43 min read

Inside the suddenly explosive world of AI safety

The Verge18h ago