OpenAI, Anthropic, Google API Flaw Lets Weaker AI Models Decode Stronger Models’ Reasoning
A newly disclosed flaw in the way OpenAI, Anthropic, and Google handle hidden AI reasoning between API calls allowed researchers to recover…
A newly disclosed flaw in the way OpenAI, Anthropic, and Google handle hidden AI reasoning between API calls allowed researchers to recover…
OpenAI has launched GPT-5.6-Cyber, a specialized cybersecurity model designed for vulnerability research, penetration testing, and incident response. Built on the GPT-5.6 Sol…
CVE-2026-15903 is a high-severity vulnerability (CVSS 8.8) in the V8 JavaScript engine, discovered by OpenAI's GPT-5.6-Cyber model. It involves an out-of-bounds read…
Security researchers created a fictitious cryptocurrency startup and hired three individuals they believe were North Korean IT operatives, as part of an…
SynthID is an invisible watermark embedded by Google in images created or edited by its AI tools. It was detected in a…
Chrome Remote Desktop is a remote access tool that allows users to control computers remotely. A suspected North Korean operative installed it…
Researchers at the University of Birmingham and the security firm Fuzzware have discovered that a malicious SIM card can execute attacker-controlled code…
Three independent research efforts presented at Black Hat USA 2026 and in subsequent disclosures have demonstrated practical attacks against passkey implementations, undermining…
New research presented at Black Hat USA 2026 reveals that CSS and HTML techniques can break out of email message boundaries to…
A recent wave of cyber attacks targeting financial services, private equity, and professional services has been attributed to the data extortion group…