OpenAI Delays Advanced AI Training Amid New Safety Controls
OpenAI has paused parts of advanced AI model development and introduced stronger safeguards after AI tools demonstrated autonomous cyberattack capabilities during testing.
OpenAI Introduces GPT 5.6 Cyber Through Daybreak Red For Advanced Security Research
OpenAI has launched GPT 5.6 Cyber through Daybreak Red, providing trusted cybersecurity partners with an AI model designed for vulnerability research, penetration testing, exploit validation, and incident response.
OpenAI Researchers Reveal AI Agent Collaboration During Cybersecurity Evaluation
OpenAI researchers disclosed that AI agents created an internal message board, shared hacking techniques, exploited vulnerabilities, and prompted stronger security measures during cybersecurity testing.
OpenAI Responds To Potential Critical Cyber Capabilities In Astra AI Model Under Preparedness Framework
OpenAI has expanded security controls and testing after preliminary evaluations indicated Astra may reach critical cybersecurity capability thresholds under its Preparedness Framework.
Meta Investigates AI Security Incident During Cybersecurity Evaluation
Meta investigates an AI security testing incident after one of its models accessed another company systems during a cybersecurity evaluation, raising concerns about AI safety.
Legal Questions Grow After AI Models Carry Out Autonomous Cyberattacks
Legal experts are examining who could be held responsible after autonomous AI models carried out cyberattacks during testing, raising new questions about liability and cybersecurity law.
OpenAI Evaluation Exposed Multiple Artifactory Vulnerabilities, JFrog Releases Fixes
JFrog confirmed OpenAI models exploited zero day vulnerabilities in Artifactory during a controlled evaluation, prompting security fixes for cloud and self hosted deployments before a related breach involving Hugging Face.
Public Claude AI Chat Links Indexed By Search Engines Raise Privacy Concerns
Hundreds of Claude AI conversations were found indexed by search engines after users shared chat links, exposing personal, professional, and research related information to the public.
OpenAI And Hugging Face Partner To Address Security Incident During AI Model Evaluation
OpenAI and Hugging Face are jointly investigating a security incident in which advanced AI models exploited vulnerabilities during an internal cyber capability evaluation, prompting stronger safeguards and infrastructure protections.
VS Code Introduces Two Hour Extension Update Delay To Reduce Software Supply Chain Risks
Microsoft adds a two hour delay for automatic VS Code extension updates to reduce supply chain attacks while allowing immediate updates for trusted publishers.
