Icon: enlarge
Clearview uses photos from Twitter, Instagram, Facebook and other services without being asked
Photo: Nick Ansell / PA Wire / dpa
Daniel Therrien has found clear words: "What Clearview does, is mass surveillance, and it is illegal," said Canada's chief data protection officer on Wednesday.
The controversial start-up from New York, which specializes in facial recognition, sees it completely differently and wants to defend itself in court if necessary.
Therrien began his investigation a year ago, along with the privacy officers of Québec, Alberta and British Columbia.
Shortly after the first New York Times article about Clearview AI was published.
From this it emerged that the company of the Australian programmer Hoan Ton-That used billions of openly accessible facial photos on Facebook, Instagram, Twitter, LinkedIn and many other websites to build a database for facial recognition.
The New York-based start-up sells law enforcement agencies (and at least temporarily to other customers) around the world access to this database.
For example, these customers upload a picture of a suspicious person and Clearview searches its database for faces that look reasonably similar.
If there are hits, the customers also receive links to the origin of the respective images and thus often information on the identity of the people depicted.
According to Therrien and his colleagues, this is a violation of Canadian law by Clearview.
There is a significant risk to individuals, "the vast majority of whom have never had or will never have anything to do with a crime."
In view of the processing of such sensitive data, the company should have obtained the consent of the data subjects, but in view of the use for such "inappropriate purposes" even that would not have been enough, according to the communication from Therrien's authority.
Clearview speaks of "unlikely" risk
Clearview argued, among other things, that the use of the data is legal even without consent, since it is public.
In addition, the advantage for law enforcement officers and national security outweighs the already "unlikely" risk for those affected.
In any case, Clearview cannot be held responsible for the authorities' mistakes.
In addition, the company has no "real" connection to Canada - the Royal Canadian Mounted Police, however, is one of their paying customers.
The privacy advocates rejected this line of argument and recommended that Clearview stop offering its services in Canada, stop collecting images of people in Canada and delete all previously collected images and the biometric data derived from them.
The company did the former last July, but the picture collection goes on and on.
Unless that changes, privacy advocates will do everything in their power to get Clearview to comply with Canadian law.
Because that is not much, they are also calling for the expansion of Canadian data protection law.
So far, you don't have the option to force Clearview to delete.
Clearview is refusing to delete the data for the time being, as Hoan Ton-That confirmed to the New York Times.
He would also like to defend himself against data protection officers in court: “This is simply about public information about who has access to it and why.
We don't want a world where only Google and a few other tech companies have access to it «.
Canada isn't the only country facing Clearview legal trouble.
A multi-million dollar class action lawsuit is pending against the company in the US state of Illinois for alleged violations of biometric data laws.
There are also studies in Australia and Great Britain.
Most recently, the Hamburg commissioner for data protection, Johannes Caspar, initiated administrative proceedings against Clearview.
Icon: The mirror
pbe
