5 Things You Need to Know About the FBI's New AI-Powered Surveillance Program
- The FBI has quietly launched a pilot program called "Project Nemesis" that uses advanced AI algorithms to scan social media platforms and public forums for "pre-crime" indicators, analyzing language patterns and behavioral triggers before any illegal action occurs.
- Privacy advocates are sounding alarms, as the program can process over 200 terabytes of data per day, including private messages shared in interest groups that are publicly accessible, raising concerns about mass surveillance and constitutional violations.
- The AI system has already flagged over 12,000 individuals in its first month of operation, with 300 referred for further investigation—though human reviewers have dismissed 98% of these as false positives, sparking debates on algorithmic bias and civil liberties.
- This initiative aligns with the FBI's recent push for "predictive policing," but internal whistleblower reports suggest the software sometimes misreads innocent phrases like "I need a shovel" or "I'll end this" as terrorism references, leading to unnecessary interviews.
- Lawmakers on both sides of the aisle are calling for a congressional hearing, arguing that without transparent oversight, the FBI risks repeating past mistakes with surveillance programs that overreach, similar to controversies around PRISM and COINTELPRO.