New Attack Against RSA

The feed covers security-relevant developments including a computationally impractical RSA signature-forgery technique against unpadded (“textbook”) RSA signatures, AI-assisted cyber abuse involving reconnaissance, exploitation, credential theft, phishing, cloud compromise, and data theft, sophisticated malicious npm packages that evade installation defenses, and research showing language models can self-jailbreak after benign reasoning training. It also discusses reverse engineering of surveillance cameras and AI-generated political disinformation. No specific CVEs are identified.