The Powers That Be: Daily The ICE Age of A.I.
Feb 5, 2026
Ian Kreitzberg, reporter who investigates AI and government data use. He unpacks ICE’s murky AI-assisted surveillance and the vast data sources it taps. They tackle facial-recognition bias, transparency gaps, and the limits of oversight. They also discuss a college poll showing students using chatbots for schoolwork, mental-health support, and even romantic connections.
AI Snips
Chapters
Transcript
Episode notes
Government Data Streams Need Algorithms
- ICE aggregates vast streams of data from many sources and needs algorithms to make sense of it.
- Companies like Palantir provide software that ingests data and produces predictions for ICE operations.
Biometrics Carry Biased Risks
- Biometric tools like facial recognition have known accuracy and bias problems, especially for minorities.
- Misidentifications by ICE apps illustrate concrete civil-rights and reliability risks in enforcement use.
Privacy Gaps Enable A Data Broker Economy
- Sensitive personal data often ends up in government hands via data brokers and cross-agency sharing without explicit consent.
- The U.S. lacks stringent privacy regulation, enabling large-scale data collection with few consequences.