OpenAI Reveals Reward Hacking Led AI Agents to Exploit Zero-Days and Breach Hugging Face
OpenAI disclosed that reward hacking was a primary driver behind an AI-powered hack of Hugging Face, which occurred during cybersecurity evaluations of…
Tag: OpenAI
OpenAI Codex CLI Vulnerable to Prompt Injection
A vulnerability chain in OpenAI Codex CLI for Windows abuses prompt injection through web.run to execute host-level commands outside the sandbox.
Amazon Kiro Prompt Injection Vulnerability Enables Data Exfiltration via Kiro Powers
Cybersecurity researchers at Mindguard have disclosed a prompt injection vulnerability in Amazon Kiro, an AI-powered agentic integrated development environment (IDE), that could…
OpenAI Bans Russian ChatGPT Accounts Tied to Influence Operation Promoting Fake Think Tank
OpenAI has banned a cluster of Russian ChatGPT accounts that used VPNs to bypass access restrictions and run an influence operation. The…
New Cryptographic Context Injection Attack Could Let Web Pages Steal Grok Chat Data
Adversa AI has disclosed a novel attack technique dubbed "Cryptographic Context Injection" that can cause xAI's Grok chatbot to exfiltrate a user's…
WIRED Report Highlights AI Safety Challenges
A WIRED report detailed OpenAI's rogue-agent hack of Hugging Face and the challenges of prioritizing safety in AI development. The report underscores…
OpenAI Pauses Frontier RL Training to Bolster AI Safety Defenses
OpenAI has announced a temporary pause in reinforcement learning (RL) training for its most advanced AI models to strengthen safety and security…
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…
Zoom Annotation Flaws Could Let Meeting Participants Hijack Attendees’ Clients
Security researchers at A Security, an Israeli-founded offensive-security startup, have disclosed three vulnerabilities in Zoom's annotation tool that could allow a meeting…
GPT-5.6-Cyber: OpenAI’s Cybersecurity-Focused AI Model
GPT-5.6-Cyber is a specialized AI model from OpenAI designed for vulnerability research, penetration testing, and incident response. It has reduced safeguards for…