ANTHROPIC'S PAPER SMELLS LIKE BULLSHIT

Summary from the TLDR Dev newsletter

Anthropic's recent report on a Chinese cyber espionage operation coordinated by Claude may have been falsified. There is a lack of verifiable information and technical details. The report claims that the AI autonomously executed a significant portion of the attack, but it fails to provide IoCs, TTPs, or evidence to support its claims.