Clearview AI Investor Pitch Reveals Photo Ingestion Rate and Expansion Plans

Big scoop for Drew Harwell at the Washington Post:

The facial recognition company Clearview AI is telling investors it is on track to have 100 billion facial photos in its database within a year, enough to ensure “almost everyone in the world will be identifiable,” according to a financial presentation from December obtained by The Washington Post.


And the company wants to expand beyond scanning faces for the police, saying in the presentation that it could monitor “gig economy” workers and is researching a number of new technologies that could identify someone based on how they walk, detect their location from a photo or scan their fingerprints from afar.

Slides Harwell posted on Twitter reveal a stunning expansion of personal recognition. For example, Clearview is exploring ways of tying identity to license plates, movements, and locations. This is a brazen theft of personal liberty undertaken by a private company with virtually no accountability or regulation.

Harwell’s story is full of outrageous details, one paragraph after another. But one must be mindful they are sourced from an investor presentation that puts the company in the best possible light for people with deep pockets. While unearned optimism is allowed, outright lies are not supposed to be included, though some of the companies’ logos apparently associated with Clearview denied to Harwell any connection.

With that in mind, Clearview says it has thousands of clients in the United States alone, even as it is increasingly banned in other countries. It says it is ingesting photos at a rate of one-and-a-half billion per month, and wants to identify pretty much everyone on Earth, though that presumably excludes children and everyone from Australia, Canada, and France. It should also be excluding images posted to major social media networks, many of which have told Clearview to cease and desist its scraping practices.

It has plenty of ideas for how it wishes to use its database and, according to the slides posted by Harwell, it also wants to license them to third parties for their own uses. How does that square with its promise to only permit law enforcement uses?

Clearview has dismissed criticism of its data collection and surveillance work by saying it is built exclusively for law enforcement and the public good. In an online “principles” pledge, the company said that it works only with government agencies and that it limits its technology to “lawful investigative processes directed at criminal conduct, or at preventing specific, substantial, and imminent threats to people’s lives or physical safety.”


In his statement to The Post, [founder Hoan Ton-That] said: “Our principles reflect the current uses of our technology. If those uses change, the principles will be updated, as needed.”

If your principles change because of financial incentives, you do not have principles. Clearview is sketchy as hell and has no place being in business.

Ton-That told The Post the document was shared with a “small group of individuals who expressed interest in the company.” It included proposals, he said, not just for its main facial-search engine but also for other business lines in which facial recognition could be useful, such as identity verification or secure-building access.

He said Clearview’s photos have “been collected in a lawful manner” from “millions of different websites” on the public Internet. A person’s “public source metadata” and “social linkage information,” he added, can be found on the websites that Clearview has linked to their facial photos.

I found the second sentence here confusing, but it seems to mean that a user of Clearview is able to see where an image in the database was sourced from. The way Ton-That phrases it makes it sound like a reassurance or something, but nothing could be further from the case. The company still makes people individually opt out from the mass surveillance machine it is hoping to grow and license if they happen to live in California or Illinois. Otherwise, Clearview says it will scrape anything it can access and it is your responsibility to remove from the web anything you do not wish to contribute to its business.

Any reasonable nation should be working hard on legislation that would prevent anything like Clearview from being used, and level crippling penalties for any of its citizens’ data found to be in its systems. That is my baseline expectation. I am not optimistic it will be achieved.