Controversial Facial Recognition Company Calls Out Clearview, Demands It Ditch Its Database Of 10 Billion Scraped Images

from the Pots,-kettles-yelling-'black'-at-each-other dept

Clearview has burned its bridges inside the facial recognition tech industry. Despite it being largely morally malleable, the industry as a whole appears to have cut ties with CEO Hoan Ton-That's startup, which relies on more than 10 billion images scraped from the web to generate a database for its customers to match faces with.

The company played it fast and loose upon rollout, handing it out to whoever seemed interested, inviting them to run searches against photos of friends and family members. The "give it a spin" invitations were handed out to government agencies as well, inviting cops to play image roulette with Clearview's ever-expanding database.

To date, the efficacy of Clearview's AI remains untested. Clearview did finally volunteer to have the National Institute of Science and Technology (NIST) examine its product for accuracy but only allowed it to run a one-to-one test -- the kind that would, for instance, allow a phone user to unlock their phone via facial recognition.

This is not the product Clearview sells. Clearview sells one-to-many matching, relying on images and other personal info scraped from the internet. This practice has alienated the internet. It has also gotten Clearview kicked out of two countries and served with multiple lawsuits. Consequently, other facial recognition tech companies are cursing Clearview both over and under their breath as the company somehow manages to remain viable in the face of months of negative press coverage.

Persona non grata status apparently applies to even other controversial facial recognition tech companies. AnyVision spent some time in the media spotlight for being used by the Israeli military to surveill Palestinians. It managed to take down Microsoft with it (albeit temporarily), exposing the tech giant's pinky-swear-rejection of enabling abusive surveillance to be the lip service it was. AnyVision doubled down on its first mistake(s) by threatening to sue news agencies that reported on this factual development.

AnyVision resurfaced a couple of years later as the company behind pervasive surveillance systems deployed by Texas public schools. AnyVision appeared to be good at what it did. It matched faces per school hot lists and let administrators know any time those faces were detected. A little too good, perhaps. The system racked up 164,000 "hits" during its seven-day test run, returning as many as 1,100 matches for a single student.

AnyVision is back. Sort of. It has rebranded as Oosto, divested itself of some of its more problematic deployments, and is now taking shots at Clearview.

Back in September, it argued that facial recognition companies selling to government agencies should offer customers a blank database -- one the end users could fill with mugshots and persons of interest. This was a clear shot across Clearview's bow. Clearview's main selling point is its scraped database of 10 billion images. AnyVision also forwarded this suggestion to a number of government bodies, including the UK's Surveillance Commissioner and the NIST.

It's always good to be wary of private companies pleading for the government to regulate them more. It often means incumbents are looking for better ways to stave off competition by helping enact rigorous guidelines that upstarts can't afford to implement. And AnyVision's suggested "fixes" for facial recognition tech obviously aims to exclude Clearview from the government market in multiple countries, making it that much easier for the rebranded company to find customers in need of controversial tech that at least won't be as controversial as Clearview.

The war of words continues, albeit behind a paywall:

Oosto (formerly Anyvision) has called out Clearview AI, stating "that biometrics should be deployed with empty databases" in a recent release, referring to the US company's practice of scraping billions of photos from social media to use in its database.

Clearview has offered a rebuttal, but until this one is scraped from its host, we won't know exactly how it defended itself. But it seems clear even from this very truncated exchange of ideas that AnyVision believes Clearview is bad and Clearview believes it is good… making it a minority of one. But one thing is clear: if Clearview gets out of the scraping business it's less likely to make enemies of governments, private citizens, and its competitors in the market. But if Clearview hasn't changed its practices following a year of caustic press coverage, it's unlikely to do so just because a competitor is waving its revamped Gadsden flag in front of any regulatory agency that will listen to it.

Filed Under: facial recognition
Companies: anyvision, clearview

Company Threatens To Sue NBC For Reporting On Its Facial Recognition Tech Being Used Against Palestinians

from the how-not-to-handle-journalists dept

Earlier this week, NBC News had quite a story about a facial recognition tech company in Israel, named AnyVision, that is being used by the Israeli military to conduct surveillance on Palestinians in the West Bank. Much of the article focuses on the fact that Microsoft invested in AnyVision, at a time when Microsoft claims it's been taking the moral high ground and unwilling to work on more nefarious uses of things like facial recognition technology. The story hits on a bunch of different points that we regularly cover at Techdirt, from misuses of facial recognition to large company hypocrisy. But we're writing about it for a different reason: the way that AnyVision's CEO reacted upon being contacted by NBC reporters:

When NBC News first approached AnyVision for an interview, CEO Eylon Etshtein denied any knowledge of "Google Ayosh," threatened to sue NBC News and said that AnyVision was the “most ethical company known to man.” He disputed that the West Bank was “occupied” and questioned the motivation of the NBC News inquiry, suggesting the reporter must have been funded by a Palestinian activist group.

As a general rule, when reporters come calling, asking for comment on a questionable thing your company may be engaged in, don't threaten to sue. Also, generally speaking, calling yourself the "most ethical company known to man" probably throws up some red flags too (as does the denial of a program that NBC had documentation about...). All that really does is tell a reporter that they're on to something.

The story itself, of course, raises all sorts of questions about how governments, law enforcement, and the military might use facial recognition -- a technology that still hasn't been proven to be particularly accurate. There's a reason why many places are banning the use of the technology by law enforcement. At the very least, any such use deserves public discussion and scrutiny -- not bogus threats of lawsuits.

Filed Under: eylon etshtein, facial recognition, israel, journalism, palestinians, surveillance, threats
Companies: anyvision, microsoft, nbc