Claude Code, Copilot and Codex all got hacked. Every attacker went for the credential, not the model.

Six teams exploited Claude Code, Copilot, Codex, and Vertex AI in nine months. Every attack hit runtime credentials that IAM ...

Anthropic's new Claude Security tool scans your codebase for flaws - and helps you decide what to fix first

It uses Opus 4.7 to scan, validate, and generate patches, helping fix dangerous flaws before they can be exploited.
The Hacker News

New Wave of DPRK Attacks Uses AI-Inserted npm Malware, Fake Firms, and RATs

Claude Opus commit added malicious npm dependency in Feb 2026, enabling crypto theft and persistent RAT access.
IEEE

A Multi-Scale Code Representation Framework for Vulnerability Detection

Abstract: With the increasing complexity of modern software systems, accurately detecting software vulnerabilities has become a critical challenge in software security. Existing representation methods ...

Facial recognition data is a key to your identity – if stolen, you can’t just change the locks

You can change a stolen password or credit card, but you can’t reset your face when your biometric data is breached.
IEEE

Unsupervised Deepfake Detection via Camera Source Clustering and Temporal-Spatial Features

Abstract: Artificial Intelligence (AI) is transforming diverse domains like cybersecurity, robotics, and healthcare but are also being misused for harmful purposes. A notable example is deep fakes ...