Clearview: scraping the barrel
Clearview AI is a US facial recognition company, which since its founding in 2017 has spent its time scraping a multitude of services on the web to find photographs of people along with their name or some detail of their identity, using them to create a huge database of more than twenty billion likenesses. If you can find a photograph of yourself on the web accompanied by your name, you are more than likely in the Clearview database (and a few others).
The idea of scraping the web to get hold of all kinds of data is a powerful one, and there are multiple rulings that accredit the procedure as legal. Clearview, however, is an example of the extent to which this type of procedure can be abused and turned into something that, although it may be legal in its origin because it is limited to collecting data that is publically available, gives rise to results that openly violate any reasonable protection of privacy. In fact, the problem with Clearview is not how it obtains its data, but what it does with it.
Facial recognition technologies are relatively mature, but by no means error-free, and even more so when applied to people from non-white ethnic groups. The use of such technologies has already proved problematic, and even more so when the tool is sold to security agencies and authoritarian governments around the world, or used as a toy for millionaires.
