A database set up by a 2020 law that allows for the use of facial recognition software to identify potential criminals has already collected 638,693 entries, a parliamentary inquiry by the liberal NEOS unveiled on Monday has found.
“We are paying a huge price for the facilitation of police investigations,” Annemarie Schlack, managing director of Amnesty International Austria has said, adding that “facial recognition threatens our rights to an extent that the benefits can’t outweigh.”
The international NGO also noted the margin for error caused by the technology, which would lead to disproportionately discriminating against marginalised groups. Interior Minister Karl Nehammer refuted these claims, stressing that each match of images is reviewed by a human being.
The software matches images from surveillance cameras with other police image databases to assist police investigations.
When the facial recognition system was first introduced, the interior ministry claimed it would only be used in regard to serious crimes. However, the Interior Ministry admitted in September 2020, that the tool has also been used to investigate demonstrators.
The issue of facial recognition is currently heavily debated at the EU level. The European Commission presented its proposal for new AI rules in April which most notably aims to raise the bar for the usage of facial recognition.
However, the EU’s two top privacy watchdogs – the European Data Protection Board (EDPB) and the European Data Protection Supervisor (EDPS) – both slammed the bill in June, calling for a general ban of any use of AI for automated recognition.