Armadin

AI models have outpaced the benchmarks designed to measure their hacking capabilities, leaving regulators and security teams unable to assess the true danger of frontier systems. Stanford's 2026 AI In…

Beijing-based AI company Z.ai has released an open-weight model, GLM-5.2, that rivals Anthropic's powerful Mythos 5 in cybersecurity tasks, raising concerns as it can be downloaded and run without saf…

Critical evaluation framework for purchasing AI-powered penetration testing tools, warning that the market is fragmented and that vendors often use the same terms to describe different capabilities. I…