Biz & Tech // Business

The person behind a privacy nightmare has a familiar face

Hoan Ton-That has invaded users’ privacy before. Now he’s asking law enforcement agencies to trust him.

Welcome back to Tech Chronicle. If you want a newsletter that won’t upload your face to the cloud but recognizes your interests, please do subscribe.

A familiar face

Hoan Ton-That? Again?

That’s what I thought when I read Kashmir Hill’s excellent piece in the New York Times about the founder of Clearview AI, a secretive New York startup that scours the internet for images of people and uses them to provide police with uncannily accurate identification of suspects based on a single photo.

Hill writes that the tool, based on a database of 3 billion images, “could end your ability to walk down the street anonymously.”

After Hill asked some police customers to test Clearview’s app with photos of herself, it did something curious. It stopped returning any results for her image, and soon after company representatives called the officers asking if they were speaking to journalists. That suggests Clearview personnel had access to law enforcement agencies’ searches, and was actively reviewing them.

I wrote about Ton-That in February 2009 (“scathingly,” Hill writes), when he was living in San Francisco, developing first Facebook and then iPhone apps. He made the news for creating ViddyHo, a website that tricked users into sharing access to their Gmail accounts — a hacking technique known as “phishing” — and then spammed their contacts on the Google Talk chat app. (The episode does not appear on Ton-That’s sanitized personal website.)

When that was shut down, he tried to resurrect it under another Web domain. The new site offered this explanation for ViddyHo’s behavior: “We had a bug in our code that would send everyone a video when they logged in.” It denied engaging in phishing.

When Hill asked Ton-That about her face being flagged by the app, he laughed and told her it was “a software bug.”

A security camera at the Olympic Stadium in London in 2012. Images captured by such cameras or uploaded to social media have proliferated online, eroding people’s privacy.
A security camera at the Olympic Stadium in London in 2012. Images captured by such cameras or uploaded to social media have proliferated online, eroding people’s privacy.
Photo: Sang Tan / Associated Press 2012

As Hill points out, with the progress of artificial intelligence and the spread of social media, it was inevitable that someone would put them together and come up with a tool like Clearview. The only thing preventing it is our own restraint. San Francisco banned use of facial recognition technology by police in May, and Oakland soon followed. The city where Ton-That got his start as a software developer won’t let his creation be put to work. Broader state or federal privacy laws may be the only comprehensive answer to inventions like Clearview.

Ton-That told Hill that the company was working on ways for people to request their information be removed from its database. It now provides an address you can email for removal requests.

You must submit a head shot and a government ID along with your request.

Is that a bug?

Or a feature?

— Owen Thomas, [email protected]

Quote of the week

“You have zero privacy anyway. Get over it.” — Scott McNealy, then CEO of Sun Microsystems, to a group of reporters in 1999

Coming up

Intel reports earnings Thursday, followed by Apple Tuesday.

What I’m reading

Teddy Schleifer reports on a California bill that could make it harder for billionaires like Jack Dorsey to hide their charitable giving. (Recode)

Carolyn Said on the latest Uber change: a test where drivers can set their own fares. (San Francisco Chronicle)

Ina Fried relays Snap CEO Evan Spiegel’s explanation of how the Snapchat app has escaped the tech backlash. (Axios)

Tech Chronicle is a thrice-weekly newsletter from Owen Thomas, The Chronicle’s business editor, and the rest of the tech team. Follow along on Twitter: @techchronicle and Instagram: @techchronicle

麻将游戏大全