9/11 launched a new era of surveillance. 25 years later, AI is making it smarter

Drive past one of the roughly 140,000 automated license plate-reading cameras scattered across the United States, and it can record your license plate, details about your car, when you passed by and the direction you were heading. That information can be stored in a searchable database available to law enforcement. 

You do not have to be suspected of a crime.

Most of the cameras are made by Flock Safety, which is why they are known as “Flock cameras.” They give police a powerful tool for finding missing people, crime suspects and stolen cars, often across multiple jurisdictions. 

But the same network also lets the police — and when access is shared, state and federal agencies — trace the movements of millions of law-abiding Americans. Enormous troves of data are collected without a warrant, without an officer at the scene and at minimal cost. 

Flock cameras, geofence warrants, facial recognition and other monitoring tools have fundamentally changed the country’s surveillance architecture.

In the past, authorities typically identified a suspect first and then gathered information about that person. Increasingly, they collect and store vast amounts of data first, then search it later for a suspect or a threat. 

This collect-first, search-later approach has roots in the frantic aftermath of the Sept. 11, 2001, terrorist attacks, when Americans were reeling from the deadliest attack in U.S. history and bracing for the possibility of another. 

“The surveillance adopted and ‘legalized’ in the wake of 9/11,” said Jennifer Lynch, general counsel at the Electronic Frontier Foundation, a nonprofit civil liberties organization, “marked a distinct pendulum swing back in the direction of increased secret government surveillance.”

And now, 25 years after 9/11, artificial intelligence is making that surveillance a lot smarter.

Post-9/11 surveillance

As Americans processed the horror of Sept. 11, Washington confronted a devastating realization: Federal agencies had pieces of information about the hijackers and their plot but failed to share them effectively, and the U.S. intelligence community failed to “connect the dots.” 

“The nation was unprepared,” the 9/11 Commission, a bipartisan group created to investigate the attacks, would later conclude.

The government’s reaction was swift. 

Just 45 days after the attacks, President George W. Bush signed the USA Patriot Act, which significantly expanded government surveillance and investigative powers and loosened restrictions on information sharing. 

The government’s post-9/11 push to gather more intelligence was starkly illustrated by the National Security Agency’s bulk telephone metadata program, which was exposed in 2013 when classified documents were leaked by former NSA contractor Edward Snowden. 

Beginning in October 2001, the NSA collected enormous amounts of telephone call records, known as metadata, so analysts could search for links to suspected terrorists. The records included Americans’ calls, even without suspicion of wrongdoing. 

The program began under secret presidential authority. In 2006, after news reports revealed some aspects of the bulk collection, the government brought it under the Foreign Intelligence Surveillance Court, which authorized it under the Patriot Act. 

These records captured the date and time of calls, their duration and the participating telephone numbers, although not the content of the conversations themselves. Telephone companies were ordered to turn the records over to the NSA on an ongoing basis. 

Congress ended the program in 2015 and replaced it with a more targeted system shortly after an appeals court ruled it exceeded the Patriot Act’s authority. 

Many of the Patriot Act provisions are still federal law, while others have expired without reauthorization. 

But the broader strategy — collect large amounts of information first, then search for threats within it — has endured, experts told Straight Arrow.

“There’s absolutely a throughline” from these post-9/11 powers to today’s surveillance, said Patrick Toomey, deputy director of the American Civil Liberties Union National Security Project.

“The government’s purchases of bulk commercial data, including sensitive data like Americans’ geolocation information,” Toomey told Straight Arrow, “shares a similar philosophy.”

From national security to local policing 

In the years after 9/11, the collect-first, search-later model increasingly moved into everyday policing. Experts said smartphones accelerated that shift by putting “tracking” devices that constantly generate, store and share personal data into everyone’s pockets. 

“There’s always been a crossover between military or national defense uses, and then local policing,” said Barry Friedman, founder and director of the Policing Project at the New York University School of Law. “And that’s what we’re seeing.”

Lynch, of the Electronic Frontier Foundation, told Straight Arrow that in the early years following Sept. 11, “mass, suspicionless surveillance” was so expensive and time-consuming that only the NSA and other well-funded federal agencies could afford it. 

But today, she said, the technologies are cheap and widely available to local police. 

Katie Kinsey, chief of staff at NYU’s Policing Project, said intelligence-sharing hubs known as “fusion centers” are another legacy of the post-9/11 era. They were created to help federal, state and local agencies share data to combat terrorism. But now, Kinsey said, they are also used for “run-of-the-mill domestic surveillance.” 

Kinsey told Straight Arrow that this level of coordination may be justified when targeting a “highly sophisticated criminal activity like a terrorist plot,” but is overkill for routine policing. 

“People might feel more comfortable making certain [privacy] trade-offs,” she said, “in the face of a crime like terrorism versus, you know, shoplifting.”

‘Gamechanger’ technology

Supporters of technologies like Flock cameras say they are invaluable law enforcement tools that can help fight crime without sacrificing privacy. 

Mitchell R. Davis III, chief of police for the Village of Hazel Crest, Illinois, just outside the south side of Chicago, said Flock cameras are a “game changer.” 

To explain why, Davis told Straight Arrow about a case from his days as a homicide detective. 

In 2000, he was searching for a missing woman when investigators eventually found her car, and then her body. The break came only because an officer happened to ticket the car and put its license plate into the police system, but that was several days too late. 

“Fast forward to today,” Davis said, “and we could have found where her car had gone during the entire route, for the most part, and possibly saved the young lady’s life.”

Davis said Flock cameras can also be critical in locating people with dementia who become lost. Situations like that, he said, “happen on a regular basis all across the country.”

Flock estimates its technology helped police locate more than 10,000 missing people in 2025, a figure the company extrapolated from a survey of nearly 700 law enforcement agencies. 

Amid growing backlash over privacy concerns and reported misuse, Flock announced new safeguards in August, including cutting its recommended data-retention period from 30 days to seven and requiring multifactor authentication. Critics, such as the ACLU, say the safeguards left loopholes for data hoarding while “treating legitimate privacy concerns as mere public relations problems.”

Flock has also said it would support “thoughtful, bipartisan legislation” that puts added privacy, transparency and accountability guardrails in place.

The company did not respond to Straight Arrow’s request for comment.

Davis, who is also the president of the National Organization of Black Law Enforcement Executives, said he understands the privacy concerns. But he said police are using these cameras to fight crime, and “not for surveillance” of ordinary residents. “We are not concerned about who drives to work down this street every day.”

He acknowledged that officers have abused the technology, but said the answer is to hold them accountable rather than abandon the tool altogether. 

In other words, he said, “don’t throw the baby out with the bathwater.” 

Rising backlash

Despite these assurances, backlash is spreading.

“Throughout the country,” said Friedman, of the NYU Policing Project, “we’re seeing nothing but alarm” over the latest technologies. He compared the current pushback to the reckoning over post-9/11 surveillance that followed the Snowden disclosures. 

“People didn’t like it then,” he said, “and they don’t like it now.” 

Flock cameras have become a focal point, drawing criticism across partisan lines. 

In August, Florida Gov. Ron DeSantis said the use of Flock cameras had gotten “out of control.”

Florida recently banned local police from placing Flock cameras on state highways and revoked previously issued permits.

“No one’s been stronger on law enforcement than me,” DeSantis said. “But I don’t want to have this become a surveillance state.”

That came after Texas Gov. Greg Abbott, also a Republican, last month ordered all state agencies to halt spending on Flock cameras, although he stopped short of calling for a statewide ban. 

The resistance extends well beyond Florida and Texas. According to a tracker maintained by the Institute for Justice, more than 200 local governments across the country have now canceled, defunded or declined to renew contracts for Flock or other automated license plate-reading cameras.

Privacy expectations 

Advances in surveillance technology and AI could reshape where Americans can expect privacy. 

For decades, courts generally held that people do not have an expectation of privacy in most public spaces. 

But NYU’s Kinsey said that principle makes less sense when technologies can track someone across a city or state, recording where they go and what they do without an officer physically present.

“Being out in public is no longer kind of the right shorthand for when you give up your privacy rights,” she said.

The Supreme Court moved further in that direction this summer in Chatrie v. United States, a closely watched case involving a “geofence” warrant.

Police investigating a 2019 credit union robbery in Virginia obtained a warrant requiring Google to provide anonymized location data for all cell phones near the scene of the crime. Investigators narrowed that pool and obtained identifying information about Okello Chatrie, who was later convicted. 

In June, the Supreme Court ruled 6-3 that police conducted a Fourth Amendment search when they obtained Chatrie’s Google location history, because he had a reasonable expectation that this data was private. The court stopped short of deciding whether the particular geofence warrant was constitutional, sending that question back to a lower court.

Friedman said he expects courts to eventually put limits on what he described as “predictive policing,” in which authorities “collect loads of information about people who haven’t been accused of anything, and then search through it to try to find some reason to think they’ve broken the law.”

Those concerns are especially acute with cellphone data, the ACLU’s Toomey said, because they can “reveal incredibly sensitive details about people’s lives and associations,” including whether someone attended a protest, visited an abortion provider or addiction counselor, or called a suicide hotline. 

AI and the future

AI could raise the stakes even higher and enable “profoundly intrusive surveillance,” Toomey said.

It can combine different kinds of data, identify patterns and automate searches across enormous databases, he said, at a speed and scale that human analysts could never match.

Kinsey said the underlying problem is that “data is kind of the currency of our lives right now,” and much of it is collected when people are suspected of no wrongdoing and without their knowledge.

Still, she said the history of surveillance provides some reason for optimism. The U.S. eventually developed legal rules for once-new technologies such as wiretaps, and lawmakers and courts can do the same again.

“We’ve done it before,” Kinsey said. “I’m cautiously optimistic that we can actually still rein in some of the risks of these technologies and keep ourselves safe.”

Toomey, however, said the need for limits is urgent. A quarter century after 9/11, he said, “the pervasive power of our government’s mass surveillance regime is clearer than ever.”

Round out your reading

• Lawyer says 57 Cornell frat members were in the group chat. Could their silence be a crime?
• Florida county discovers mysterious Flock cameras with no obvious owner.
• Off Script: They won their primaries on one message. Now it’s disappearing.
• Some question whether a popular affordable housing policy is making the nation’s home shortage worse.
• Kelvin wave incoming: Why the US should care.