Brazil’s Federal Police are planning to use facial recognition to identify the perpetrators of attacks against the country’s federal agencies and institutions in Brasilia this past Sunday.
A government announcement specifies the use of biometrics “to identify those responsible for acts of vandalism,” according to a Google translation, at the Planalto Palace, National Congress and Federal Supreme Court.
Expect facial recognition and social media posts to give up the identity of many attackers in the days ahead. Another source of information for police, perhaps the main one, will be the digital accounts and devices of those who participated in the attacks.
This is made clear in an examination of how the similar attack on the U.S. capital building in January of 2022 has been investigated, published by IEEE Spectrum.
The analysis shows the use of facial recognition to identify rioters, but only as the initial lead in about 25 cases.
Most of those charged have been identified through different means; most notably, through digital records found online or on their personal devices.
Clearview AI has crafted a webpage showing how its biometric technology was used in investigations, and led to the identification of people not found in typical law enforcement databases since they were first-time offenders.
The Spectrum article suggests that in more cases, over 34 percent in fact, geofencing yielded the initial clue to the rioter’s identity. Image recognition provided that first clue about half as often, and not always in the form of a biometric match. Facebook itself seems to have turned up a greater number of productive leads.
Brazil’s government does not specify an intention to use digital forensics, but as the IEEE article shows, for many of those who have left evidence of their crimes online, or even their own smartphones, it is probably too late to hide the evidence.