A Simple Phrase Defeats GPT5 Security

In this episode of Cybersecurity Today, host Jim Love discusses recent developments in cybersecurity, including a method to bypass GPT5 model safeguards, malware issues in the Google Play Store, NIST’s new AI-specific security controls, and a cyber attack that led to a government shutdown in Nevada. The episode also covers a CRM-related breach linked to the Shiny Hunters collective, who used OAuth tokens to gain unauthorized access. Key takeaways emphasize the need for stronger security frameworks and vigilance against evolving cyber threats.

00:00 Introduction and Overview
00:27 Exploiting GPT-5: A Simple Prompt Attack
02:20 Google Play Store’s Malware Struggles
04:11 NIST’s New AI Security Controls
06:06 Nevada Government Cyber Attack
08:23 Shiny Hunters’ CRM Breach
10:41 Conclusion and Contact Information

This article has been indexed from Cybersecurity Today

Read the original article: