AI Security: Key Challenges in the AI Era
During the 15th Information Security Day seminar, expert Park Chan-am highlighted critical AI security issues, emphasizing the need for advanced threat management and internal software security.
TechHorizon.news is an independent news platform covering cybersecurity, artificial intelligence, and digital transformation, with a focus on risk, regulation, and executive decision-making across Europe and the wider region.
During the 15th Information Security Day seminar, expert Park Chan-am highlighted critical AI security issues, emphasizing the need for advanced threat management and internal software security.
The GhostApproval vulnerability affects several AI coding assistants, allowing bypass of safety controls and potential code execution on developer machines due to symlink exploitation.
A cybercriminal group named Lurking Lizard has been using fake 7-Zip installers to turn victim computers into proxy servers. Investigations reveal a vast operation dating back to 2022 and involving over 230 domains.
OpenMatter Network has joined the HOL Partner Program to develop standards for secure autonomous AI systems. This collaboration emphasizes the need for verification over trust in AI interactions.