REGULATION Anthropic to Brief Global Financial Regulators on Cyber Flaws Claude Mythos Found 8/10 3 min read 1 month ago
BENCHMARKS Claude Mythos Tops ExploitBench, Develops V8 Browser Exploits at 12x GPT-5.5 Cost 8/10 3 min read 1 month ago
BENCHMARKS Claude Mythos Preview Becomes First AI Model to Clear All UK AISI Cyberattack Simulations 7/10 3 min read 2 months ago
REGULATION OpenAI Joins Anthropic in Restricting Powerful Cybersecurity AI Access 8/10 4 min read 3 months ago
RESEARCH University of Florida Researchers Achieve 99 Percent AI Jailbreak Rate Using Internal Model Architecture Attacks 8/10 2 min read 3 months ago