
Google Has Deployed A New AI Spam Detector Called SAFE via @sejournal, @martinibuster
- ●Google has introduced a new system called Scaled Abuse Forensics Examiner (SAFE) aimed at detecting AI-generated spam content that violates platform policies.
- ●SAFE identifies 'spirit of policy' violations by leveraging few-shot-trained LLMs, aiming to catch content that doesn't match specific rules but still undermines platform guidelines.
- ●The system improves forensic investigations by reducing the time needed to identify novel synthetic threats through automated processes and specialized AI agents.



