Latest security news
- A $1,400 experiment in AI security auditing outperformed OpenAI’s Codex Security
- Residential proxy SDKs are hiding in LG and Samsung smart TV apps
- Free, no-signup World Cup streams serve scams instead of football
- Only 7% of companies are ready for the AI agents they deployed
- GentleKiller Framework Disables Victims' Security Software
- Unpatchable BootROM Flaw Impacts Apple A12, A13 Chips
- Microsoft Attributes Mastra AI Supply Chain Attack to North Korea
- Klue Breach Enables Hackers to Compromise Cybersecurity Firms via OAuth Tokens
- UK Information Commissioner Resigns After Workplace Investigation
- NCSC Urges Fortinet Customers to Tackle FortiBleed Fallout
- Who pays when you gate cyber-capable AI models?
- Agent Beacon: Open-source telemetry layer for AI agents
- Encrypted DNS still tells an eavesdropper where to look
- Product showcase: Avira Security for iOS blends security, privacy, and device optimization
- Hundreds of AI-powered iOS apps found exposing credentials
- AWS Unveils 'Continuum,' an AI-Powered Vulnerability Management Platform
- Operation Endgame Disrupts Malware Network Linked to Major Ransomware Gang
- Confidence Lacks in Threat Detection Across Non-Email Channels like Slack and Teams
- Fake GitHub Stars and AI Videos Mask a Crypto Clipper
- ICO Cautions Healthcare Worker After Princess of Wales Incident
- LATAM Infrastructure Hit by Fortinet and Ivanti Exploits
- Hostile States Behind 75% of Cyber-Attacks on UK Critical Infrastructure, NCSC Warns
- Cybercrime Surges in APAC as Digitalization Takes Hold
- North Korean Hiring Fraud Runs on AI and US Laptop Farms
A research team has built a system that teaches AI agents to hunt for software bugs by writing the audit method down as plain text. The system, called EVOHUNT, keeps the underlying AI model fixed and improves only an external “playbook” that tells the agent how to work. One result stands out for anyone buying security tools. An open-source model running an evolved playbook found real vulnerabilities at a higher rate than OpenAI’s commercial Codex …
(C) Do-Know.com (http://do-know.com/). Do not copy without permission from info at do-know.com.