Thought Leadership
Jailbreak
Researchers at UNSW Sydney have induced language models to imitate drunken behavior. The models subsequently revealed secrets and answered requests they were designed to block.
Fake Cloudflare Check
Users were directed through Google ads to a malicious Custom GPT that sent them to a ClickFix page delivering a remote access trojan.
Penetration Testing as a Cover
A Chinese-speaking hacker used AI models from Anthropic, DeepSeek, and Moonshot to steal more than 600,000 credit card records within five days.
AI-Powered Extortion
Anthropic uncovered a case in 2025 in which a hacker used Claude Code to automate almost the entire operation of an extortion campaign.
AI Agent Bypassed Access Controls
An autonomous AI agent developed by US-based OpenAI hacked a government portal in Australia.
Seven-Day Deadline
The ShinyHunters hacking group claims to have breached FBI systems and stolen sensitive data belonging to agents and job applicants.
Complete Takeover Packages
According to an analysis by NordStellar, SIM-swapping services are available on the dark web for as little as $300 USD, often bundled with stolen identity data.
ChatGPT Account Access
Security researchers at Hacktron AI used Anthropic’s Claude models to chain two vulnerabilities and gain access to OpenAI’s internal systems.
Fake payment error email
A zero-day vulnerability dubbed StyleSmuggler in all versions of Magento as well as Adobe Commerce is already being actively exploited for attacks.
Over 5,000 stolen credentials
A framework operated as Phishing-as-a-Service called BigBear 2.0 has bypassed multi-factor authentication at 258 organizations, stealing more than 5,000 Microsoft 365 credentials in the process.
Most Read Articles
28. September 2026
27. September 2026
13. September 2026
11. September 2026
10. September 2026