Flock and oversight: Trust is harder to build after the fact

BY BRENT MCCONNELL

In 2008, I was the U.S. national security lead for technology transfer to China. Long before the Beijing Olympics opened, many requests for advanced technology export licenses for cameras built to watch crowds and identify faces landed on my desk. The requests said they were for Olympic security. Fair enough, I thought. A major international event has real security risks. Then the Olympics ended, and the requests kept coming.

If the need was really about the Games, the orders should have stopped once the Olympic security mission ended. They didn’t. I recommended denial. The Olympics were over, but the requests were still being justified as Olympic security. I believed those cameras would become a powerful tool for China’s government to control and surveil its own citizens. I still think that call was right. I lost anyway. Commercial interests wanted the sale, and the sale happened.

That same category of technology – networked cameras, facial recognition and searchable surveillance data – later became part of China’s broader system of internal control. China’s social credit system is a mix of blacklists, ratings, court records, financial records, traffic enforcement, local experiments and government databases. Citizens with low scores can be denied employment, airline tickets or college entrance. In China, these tools operate inside an authoritarian political system. There is no independent court in the way Americans understand it. There is no free press able to investigate freely. There is no election that can remove the officials who run the system. That is the point. The camera is not what makes the system authoritarian. The absence of anyone who can effectively say no is what makes it dangerous. A democracy cannot adopt the tool and ignore that lesson.

I have earned my skepticism of surveillance cameras honestly. But technology does not always play out the same way. Some technologies that worried people when they first appeared went on to do very good things.

In 1888, Kodak marketed its first portable camera. Newspapers reacted with genuine alarm. People even used the phrase “Kodak fiends” for those who snapped pictures of strangers in public places. Boston lawyers Samuel Warren and Louis Brandeis were bothered enough that in 1890 they wrote the essay credited with introducing the American legal idea of a right to privacy.

The new camera scared people because it allowed strangers to capture and keep someone’s image without permission. That was a real concern. But over time, personal photography spread around the world. Kodak put cameras in ordinary people’s hands. Laws, court decisions and social habits slowly caught up. Today, almost everyone carries a camera in a phone.

DNA evidence took a different path to a similar place. When forensic DNA testing entered courtrooms in the mid-1980s, civil liberties lawyers had good reasons to worry. Would juries treat the science as perfect? Would DNA databases become permanent records of anyone ever arrested? 

Some of those concerns were fair. But the field developed rules. Labs became accredited. Testing methods improved. Courts learned to question bad science before a jury heard it. DNA evidence now both convicts guilty people and clears innocent people who were wrongly convicted. Public concern eased because rules and oversight developed around the technology.

Which brings us to Flock.

A case close to home shows the upside. In December 2025, someone stole a credit card from a car parked at Boca Bay Pass Club. Lee County Sheriff’s Office reviewed surveillance videos, identified the thief’s car, worked through Gasparilla Island Bridge Authority toll booth footage and used license plate reader data to tie the car to a plate and a name.

That is exactly the kind of case where most people want technology to work. A crime happened. Investigators followed the evidence. 

While Flock does not operate exactly the same as the facial recognition technology we sold to China, it still captures private citizen information, and that information can be used against those citizens. While the two camera technologies are slightly different, the governance problem is similar; the capability is there, but is the accountability and structure to protect us from abuse of this technology also there?

Flock rolled out quickly, one city contract, neighborhood, business and police agency at a time. Many people did not understand how broad the network had become or how far the data could travel. Some cities later discovered that license plate data from their cameras had been accessible to outside agencies without their knowledge or approval. Communities around the country are now pausing, canceling or reviewing contracts because the public trust arrived after the technology, not before it.

The same cross-agency access that can help solve a Boca Grande theft can also become a surveillance network if nobody controls it carefully. The technology is useful but that does not make it harmless. The danger is not the camera alone. It is a camera connected to a database, connected to other agencies, connected to few rules, connected to little public understanding.

The camera and DNA testing both eventually developed enough oversight to earn public trust, though by different routes. Flock is now finding out, one canceled contract and city council meeting at a time, whether it can build that same trust after the fact.

It may still be possible. But trust is easier to build before people feel watched than after they discover they already were.

McConnell of Cape Haze is a former Pentagon lead for weapons security. Email AITechThoughts@gmail.com