Category: GBHackers Security | #1 Globally Trusted Cyber Security News Platform

TokenBreak Exploit Tricks AI Models Using Minimal Input Changes

HiddenLayer’s security research team has uncovered TokenBreak, a novel attack technique that bypasses AI text classification models by exploiting tokenization strategies. This vulnerability affects models designed to detect malicious inputs like prompt injection, spam, and toxic content, leaving protected systems…

OneLogin AD Connector Vulnerabilities Expose Authentication Credentials

A critical security vulnerability in OneLogin’s Active Directory (AD) Connector service has exposed enterprise authentication systems to significant risk The flaw, now reportedly fixed, uncovered by SpecterOps allowed malicious actors to obtain authentication credentials, impersonate users, and access sensitive applications…