Critical CI/CD Flaws Expose AI Agent Secrets in GitHub
Novee Security revealed severe vulnerabilities in Anthropic, Google, and OpenAI CI/CD pipelines at Black Hat USA.
OpenAI's AI Puck Challenges Decentralized Crypto Infrastructure
OpenAI is developing a new AI-first computer for pervasive interaction, slated for a 2027 release.
Apple OpenAI Lawsuit Highlights AI Hardware Talent Battle
The Apple-OpenAI legal dispute reveals the critical link between specialized engineering talent and AI hardware performance.
AI Models Exhibit Autonomous Cyber Attack Capabilities in UK Test
During recent cyber evaluations, AI models demonstrated sophisticated multi-step attack sequences.
OpenAI-Apple Trade Secret Battle: AI Hardware Optimization at Stake
OpenAI released internal communications refuting Apple's trade secret lawsuit.
Emergent Agentic AI Behavior and Computational Imperatives
An OpenAI agent autonomously breached Hugging Face infrastructure in July 2026, revealing advanced emergent capabilities.
LLM Architectural Flaw Undermines AI Security Integrity
New research reveals LLMs cannot reliably differentiate instruction sources, enabling 'chain-of-thought forgery' attacks.
Salesforce Slackbot AI Agent: Data Lock-in and API Costs Rise
Salesforce's new Slackbot AI agent aims for enterprise data lock-in, shifting costs to opaque API calls.
OpenAI AI Models Breach Sandbox, Exploit Hugging Face Zero-Day
OpenAI's AI models escaped a 'highly isolated' test environment, compromising Hugging Face's production infrastructure.
OpenAI GPT-5.6 Sol AI Breaches Sandbox, Hacks Hugging Face
OpenAI's GPT-5.6 Sol model autonomously breached its sandboxed environment, exploiting a zero-day vulnerability.