AI

Claude AI Finds SAML Security Flaws That Can Let Attackers Take Over Accounts

Security researchers have used Anthropic’s Claude AI to uncover serious flaws in Security Assertion Markup Language (SAML) implementations that could…

4 weeks ago

Irregular Pushes Stronger Containment Standards for Secure Frontier AI Cyber Evaluations

Irregular has detailed an investigation into an AI cyber-evaluation incident in which internet access unintentionally allowed models to interact with…

4 weeks ago

Anthropic Launches New /design Skill for Claude Code UI Workflows

Anthropic has introduced a new /design skill for Claude Code, expanding its developer-focused AI platform into user interface design workflows.…

4 weeks ago

Z.ai Unveils GLM-5.3 with Major Enhancements for Coding and Cybersecurity

Z.ai has released GLM-5.3, a new AI model built to handle complex coding, long-running agent tasks, and cybersecurity analysis. The…

4 weeks ago

New MessiahGPT AI Model Fueling Automated Ransomware and Phishing Attacks

A criminal AI service called MessiahGPT is being marketed on BreachForums as an uncensored platform for creating ransomware, phishing kits,…

1 month ago

Microsoft Begins to Merge Consumer and Enterprise Copilot Apps to Make New Super App

Microsoft is moving closer to a unified Copilot experience by merging key elements of its consumer AI assistant with the…

1 month ago

Blacklight Toolkit Finds Codex, Claude Code, and Cursor Artifacts Exposing Tokens and Session Data

SpecterOps has released Blacklight, an open-source toolkit that identifies local artifacts from AI coding agents like Codex, Claude Code, Cursor,…

1 month ago

New ‘Ghostjacking’ Attack Lets Hackers Hijack AI Agents to Run Their Code on Developer Machines

A new AI-agent attack technique called “Ghostjacking,” can trick coding agents into running attacker-controlled commands, changing cloud settings, stealing credentials,…

1 month ago

Anthropic to Add Invisible Watermarks to All Claude AI Outputs

Anthropic has announced plans to add invisible watermarks and signed metadata to content created by Claude AI. The move follows…

1 month ago

Claude Opus 5 Cuts Indirect Prompt Injection Attack Success to 2% in New Benchmark Analysis

Anthropic’s Claude Opus 5 has recorded the lowest indirect prompt injection attack success rate in Gray Swan’s latest benchmark, according…

1 month ago