Discreetly, Clearview AI is trying to attract new investors in order to build an unprecedented database of 100 billion photos, used by its facial recognition tool. Objective: to expand internationally.
According to the documents from a financial presentation held last December, obtained by the washington postcompany Clearview AI, whose facial recognition technologies are controversial, is reportedly considering collecting enough photos to ensure it has almost every human being in its database.
“Clearview AI tells investors it is on track to collect 100 billion portraits in its database within a year, enough to ensure ‘almost everyone’ will be identifiable globally“, writes the American daily. Still according to the washington post, Clearview AI reportedly already has 10 billion images in its database, which is currently growing at the breakneck pace of 1.5 billion new portraits each month. The company has even quantified its financial needs to reach 100 billion photos, which it estimates at 50 million dollars.
The plan presented to investors is as clear as it is terrifying: finish composing this international base to improve its facial recognition solution – or even create new products of this kind – in order to launch its marketing in new countries. Local sales teams could be recruited in these markets, and lobbying investments will be made there, in order to put pressure on public authorities to generalize the use of facial recognition.
Among the tracks of new applications on which board Clearview AI appears the identification of people via the analysis of their step (a process already tested in China which makes it possible to identify the people of profile or back) as well as the taking of remote fingerprint.
Decried methods
As a reminder, Clearview AI has started to compose its database by sucking up photos and videos publicly available on social networks. Facebook, Google, Twitter and even YouTube have all indicated that this operation constitutes a violation of their rules of use and have ordered Clearview AI to cease such practices and to erase the data thus obtained. What Clearview AI refuses to do. Hoan Ton-That, boss of the company, indicates that this data is protected by the first amendment. He adds that the database is constituted legally, via millions of different Internet sites, and compares this to the indexing of sites by Google.
Finally, he indicates that Clearview AI has chosen not to market its solution to the general public, selling it only to law enforcement to help them in their investigations in order to fight against crime. However, its application has ended up in the hands of certain celebrities, and not only government agents. It is also not certain that the economic model of Clearview AI will not evolve in the future, which – according to Hoan Ton-That – would give rise to an update of its public commitments.
Among the interested platforms, we can list Airbnb, Lyft or even Uber, all three of which have already approached Clearview AI. They would seek, by using facial recognition, to improve the security of their services. For the time being, these possible collaborations remain hypothetical, no project of this kind having officially been approached by Clearview AI. On the other hand, Hoan Ton-That is proud to remind anyone who will listen that its technology was used to identify the rioters of January 6, 2021 in Washington DC, following the capture of the United States Capitol.