While the AI itself wasn't designed for harm, the underlying method sparks worry regarding AI agents with extensive system permissions. Image Source: thinkhubstudio/Shutterstock A recently discovered security flaw has led…