FBI director touts expanded AI use for threat cases
FBI Director Kash Patel said in a weekend post on X that the bureau is ramping up its use of artificial intelligence, describing AI as a 'key component' of its strategy to stay ahead of domestic and international threats and announcing a formal AI initiative to support national security investigators and analysts. Patel also revealed that outgoing Deputy Director Dan Bongino is leading a new internal technology working group to ensure FBI tools evolve with the mission, while the bureau highlights existing AI uses such as vehicle recognition and voice and video analytics.
📌 Key Facts
- FBI Director Kash Patel stated on X that artificial intelligence is a 'key component' of the bureau’s strategy for an 'always changing threat environment' at home and abroad.
- Patel said the FBI is developing an AI project specifically to assist investigators and analysts working in the national security space.
- He announced a 'technology working group' led by outgoing Deputy Director Dan Bongino to guide the bureau’s technology evolution.
- The FBI already uses AI for vehicle recognition, voice-language identification, speech-to-text analysis and video analytics, according to its website.
- A bureau spokesperson told Fox News Digital it had nothing to add beyond Patel’s public post, underscoring that this is an official, on-the-record initiative.
📊 Relevant Data
Facial recognition algorithms exhibit false match rates up to 100 times higher for individuals from West Africa (proxy for Black/African groups) compared to those from Eastern Europe (proxy for White groups), with a specific factor of 720 for elderly Nigerian women versus middle-aged Polish men.
Face Recognition Vendor Test (FRVT) Part 8: Summarizing Demographic Differentials — National Institute of Standards and Technology (NIST)
False non-match rates in facial recognition are 2-10 times higher for West African groups (proxy for Black/African) compared to Eastern European groups (proxy for White), with rates around 0.04 for Nigerian females versus 0.005 for Eastern European males.
Face Recognition Vendor Test (FRVT) Part 8: Summarizing Demographic Differentials — National Institute of Standards and Technology (NIST)
Deployment of facial recognition technology by police is associated with increases in Black arrest rates and decreases in White arrest rates across U.S. cities.
Facial recognition systems in policing and racial disparities in arrests — Government Information Quarterly (ScienceDirect)
Top-performing facial recognition algorithms have overall false non-match rates of 0.002 to 0.005, indicating high accuracy in controlled settings, but with demographic variations.
Face Recognition Vendor Test (FRVT) Part 8: Summarizing Demographic Differentials — National Institute of Standards and Technology (NIST)