Imagine an America where multiple police officers and security guards stand watch on every block, in every park, in every store, and in every other public space around the clock. Imagine these officers watching us constantly, not only scrutinizing our every move for signs of “suspicious” behavior, but also noting many details about us — such as what we’re wearing and carrying, who we’re with and what our relationship appears to be with them — and recording those details in a searchable database.

That’s not going to happen. Nobody wants to pay for human officers to stand watch in places where almost nothing ever happens. But we are moving toward a future where we might end up with an automated machine equivalent. Video analytics, the technology to make that possible, has been developing for many years. Now, the same generative artificial intelligence (AI) techniques that have revolutionized large language models (LLMs) like ChatGPT are in the process of creating a new, more powerful generation of this technology that could super-charge video surveillance.

In 2019, the ACLU published a report on how video analytics makes it possible for machines to not jut record oceans of video, but to “watch” that video — in the sense that they’re able to analyze, in real time, what’s happening in a video feed — and send alarms to humans when certain conditions are met. Video data, which was formerly very difficult to search and analyze, has also become increasingly searchable through queries such as “find me a male wearing a purple shirt and carrying a violin case” that (like face recognition) can now be run across vast amounts of video data. Since then, video analytics technologies have become widely available, with most commercial surveillance cameras including some form of the technology built in.

The older generation of video analytics, however, is limited to detecting a narrow set of objects on which it has been laboriously trained, and often performs poorly — “sold and marketed way beyond real-life performance,” as one industry player put it. But today the revolutionary advances in large language models are in the process of spawning a new generation of the technology. While language models, per their name, are mostly focused on text, the techniques and advances that led to those models’ breakthrough success are spilling into machine vision as well — specifically programs dubbed “Vision Language Models” (VLMs) that can understand both visual and natural-language textual inputs. In computer science terms, these new machine vision programs are based on the same technology as language models, called transformers, as opposed to classic machine vision work, which is based on a technology called convolutional neural networks (CNNs). While both technologies continue to be used and sometimes combined, and video technologies are still evolving fast, this appears to be a big change.

The advent of vision language models will have three important effects.

VLMs are able to generalize much better than the older, CNN-based video analytics programs because they combine image recognition with the general world-knowledge that large language models gain as part of their training on all the Internet’s textual data.

In the older form of machine vision, for example, a CNN might be shown millions of pictures of horses and elephants and thus laboriously learn to identify and distinguish them. An LVM, on the other hand, might be able to find a zebra in a video even if it had never seen a photo of a zebra before, simply by leveraging its world knowledge (that a zebra is like a horse with stripes). Instead of being limited to a closed set of predefined things, VLMs are able to recognize an enormous variety of objects, events, and contexts without being specifically trained on each of them. VLMs also appear to be much better at contextual and holistic understandings of scenes.

The CEO of security analytics company Ambient.ai declared this shift “the most significant technology evolution in the history of video analytics ever,” saying “it solves all the problems that kept traditional analytics from getting the last miles to large-scale adoption.” Logan Kilpatrick, manager of Google DeepMind, told a podcaster that

my guess is that as [VLM-based] vision becomes more and more prominent, we’re going to see [startups] go after all of these eco-systems and industries where they’re using domain-specific vision models and not using a general purpose model. And… you unlock all these use-cases which those models are just not actually capable of doing; they’re very very rigid and can’t be fault tolerant in a lot of those cases.

Anybody can gain a sense of the power of the new models using this site created by a former Google engineer to teach people just how much information can be extracted from their photos by AI. Or by going directly to a site like Google’s AI Studio and play around with uploading photos and videos. In addition to detailed descriptions of objects and people, the models can make observations on things like emotional state and even social class.

In December the technologist Simon Willison calculated that to analyze all of the 68,000 images in his personal photo library using the Google Gemini model would cost $1.68. It’s also possible to stream videos to models like Gemini and have them analyze the contents, which appears to cost roughly 10 cents per hour of video. Such low costs mean that as the technology is refined, and as understanding of these capabilities spreads, it will not be confined to Google and a handful of other AI developers. The technology will become easily accessible to a broad variety of security companies and find its way into the products that are used to monitor us across a wide variety of contexts, from private spaces like stores and shopping malls, to those public spaces where police departments have deployed surveillance cameras.

As with LLMs, the models may also increasingly become possible to run locally, without having to connect to the servers of, and share data with, OpenAI, Google, or other big companies. It’s good if AI technologies are democratized rather than being controlled by big players, but that also means that guardrails — such as those we recommended in our report — are going to become vital as various parties, well-intentioned and not, deploy them.

Instead of being confined to precisely worded menus or tags of objects and behaviors that a model has been trained to recognize, users can just issue commands using everyday speech, such as “text me if the dog jumps on the couch,” “let me know if any kids walk on my lawn,” or troublingly, “alert me if a Black man enters the neighborhood” or “if someone is behaving suspiciously.”

It’s important to keep in mind that like large language models, vision language models are unreliable. The surveillance industry analysis firm IPVM tested one security company’s new LVM-powered product and observed that it “returned some results that were incredibly impressive but also some results that were incredibly bad.” A group of academic and industry experts explained in a recent paper that

connecting language to vision is not completely solved. For example, most models struggle to understand spatial relationships or count… They often ignore some part of the input prompt [and] can also hallucinate and produce content that is neither required nor relevant. As a consequence, developing reliable models is still a very active area of research.

As with face recognition (which is actually a subset of video analytics), there are reasons to worry about this technology when it works poorly — and other reasons to worry when it works well. If LVMs remain unreliable, but just reliable enough that people depend on them and don’t double-check that results are accurate, that could lead to false accusations and other injustices in security contexts. But to the extent it becomes more intelligent, that will also allow for more and richer information to be collected about people, and for people to be scrutinized, monitored, and subjectively judged in more and more contexts.

In the end, nobody knows how capable this technology will become or how quickly. But policymakers need to know that advancing AI means surveillance cameras no longer the classic cameras of yesterday that do nothing more than record. Already we’re seeing AI used for monitoring in an increasing number of contexts, including vehicle driver monitoring, workplace monitoring, gun detection, and the enforcement of rules. If we let it happen, we can expect that nearly every rule, regulation, law, and employer dictate that can be enforced through visual monitoring of human beings will become subject to these unblinking and increasingly intelligent yet unreliable artificial eyes.

Earlier this month, recent graduate, activist, soon-to-be father, and legal permanent resident Mahmoud Khalil, was arrested and detained in direct retaliation for his advocacy for Palestinian rights at Columbia University. Later, the Department of Homeland Security (DHS) transferred him to a Louisiana detention facility 1,400 miles away from his home and his family.

Following his illegal arrest, a team of lawyers, including Amy Greer from Dratel & Lewis, the Center for Constitutional Rights, and CLEAR secured a court order to block his deportation. Since then, the New York Civil Liberties Union (NYCLU), American Civil Liberties Union (ACLU), and Alina Das, co-director of the Immigrant Rights Clinic at New York University (NYU) School of Law have joined his legal team. His lawyers are arguing that his detention violates his constitutional rights, including free speech and due process, and goes beyond the government’s legal authority.

Below, read a letter Khalil dictated over the phone from Immigrations and Customs (ICE) detention in Louisiana – his first public statement since his arrest.

---

Letter from a Palestinian Political Prisoner in Louisiana March 18, 2025

My name is Mahmoud Khalil and I am a political prisoner. I am writing to you from a detention facility in Louisiana where I wake to cold mornings and spend long days bearing witness to the quiet injustices underway against a great many people precluded from the protections of the law.

Who has the right to have rights? It is certainly not the humans crowded into the cells here. It isn’t the Senegalese man I met who has been deprived of his liberty for a year, his legal situation in limbo and his family an ocean away. It isn’t the 21-year-old detainee I met, who stepped foot in this country at age nine, only to be deported without so much as a hearing.

Justice escapes the contours of this nation’s immigration facilities.

On March 8, I was taken by DHS agents who refused to provide a warrant, and accosted my wife and me as we returned from dinner. By now, the footage of that night has been made public. Before I knew what was happening, agents handcuffed and forced me into an unmarked car. At that moment, my only concern was for Noor’s safety. I had no idea if she would be taken too, since the agents had threatened to arrest her for not leaving my side. DHS would not tell me anything for hours — I did not know the cause of my arrest or if I was facing immediate deportation. At 26 Federal Plaza, I slept on the cold floor. In the early morning hours, agents transported me to another facility in Elizabeth, New Jersey. There, I slept on the ground and was refused a blanket despite my request.

My arrest was a direct consequence of exercising my right to free speech as I advocated for a free Palestine and an end to the genocide in Gaza, which resumed in full force Monday night. With January’s ceasefire now broken, parents in Gaza are once again cradling too-small shrouds, and families are forced to weigh starvation and displacement against bombs. It is our moral imperative to persist in the struggle for their complete freedom.

I was born in a Palestinian refugee camp in Syria to a family which has been displaced from their land since the 1948 Nakba. I spent my youth in proximity to yet distant from my homeland. But being Palestinian is an experience that transcends borders. I see in my circumstances similarities to Israel’s use of administrative detention — imprisonment without trial or charge — to strip Palestinians of their rights. I think of our friend Omar Khatib, who was incarcerated without charge or trial by Israel as he returned home from travel. I think of Gaza hospital director and pediatrician Dr. Hussam Abu Safiya, who was taken captive by the Israeli military on December 27 and remains in an Israeli torture camp today. For Palestinians, imprisonment without due process is commonplace.

I have always believed that my duty is not only to liberate myself from the oppressor, but also to liberate my oppressors from their hatred and fear. My unjust detention is indicative of the anti-Palestinian racism that both the Biden and Trump administrations have demonstrated over the past 16 months as the U.S. has continued to supply Israel with weapons to kill Palestinians and prevented international intervention. For decades, anti-Palestinian racism has driven efforts to expand U.S. laws and practices that are used to violently repress Palestinians, Arab Americans, and other communities. That is precisely why I am being targeted.

While I await legal decisions that hold the futures of my wife and child in the balance, those who enabled my targeting remain comfortably at Columbia University. Presidents Shafik, Armstrong, and Dean Yarhi-Milo laid the groundwork for the U.S. government to target me by arbitrarily disciplining pro-Palestinian students and allowing viral doxing campaigns — based on racism and disinformation — to go unchecked.

Columbia targeted me for my activism, creating a new authoritarian disciplinary office to bypass due process and silence students criticizing Israel. Columbia surrendered to federal pressure by disclosing student records to Congress and yielding to the Trump administration's latest threats. My arrest, the expulsion or suspension of at least 22 Columbia students — some stripped of their B.A. degrees just weeks before graduation — and the expulsion of SWC President Grant Miner on the eve of contract negotiations, are clear examples.

If anything, my detention is a testament to the strength of the student movement in shifting public opinion toward Palestinian liberation. Students have long been at the forefront of change — leading the charge against the Vietnam War, standing on the frontlines of the civil rights movement, and driving the struggle against apartheid in South Africa. Today, too, even if the public has yet to fully grasp it, it is students who steer us toward truth and justice.

The Trump administration is targeting me as part of a broader strategy to suppress dissent. Visa-holders, green-card carriers, and citizens alike will all be targeted for their political beliefs. In the weeks ahead, students, advocates, and elected officials must unite to defend the right to protest for Palestine. At stake are not just our voices, but the fundamental civil liberties of all.

Knowing fully that this moment transcends my individual circumstances, I hope nonetheless to be free to witness the birth of my first-born child.

It was another busy week at the ACLU of Maine: We gathered with the community at SPACE Gallery, outlined our Robbins victory and next steps, and testified in Augusta on one piece of the puzzle to end the state's Sixth Amendment crisis and uphold the right to counsel.

Come As You Are Community Resistance Event at SPACE Gallery 

We had an amazing time at the “Come as You Are” community resistance event at Portland’s SPACE Gallery on Tuesday, hosted by the Maine Writers & Publishers Alliance. The evening featured readings by several Maine writers, including our own Molly Curren Rowles, who focused on perseverance and resistance during a tumultuous and challenging time. Keep an eye on our events page to see where we’ll be in the coming weeks and months! 

 

Read Molly's Piece

 

Spreading the Word About Our Recent Court Victory and Next Steps

Our chief counsel, Zach Heiden, wrote a piece in the Maine Morning Star detailing our recent victory in Robbins v. State of Maine, and the long journey that got us here. The piece puts into perspective the outsized impact that failing the Sixth Amendment has on the people of Maine and outlines the next steps in the state’s attempt to fix this ongoing constitutional crisis. Read the piece here, and check out our dashboard that tracks the number of people in Maine without an attorney starting in November of 2023.

 

MAINE MORNING STAR

ACLU of Maine Testimony in Support of Expanding Our Public Defense System

We had a busy week in Augusta, testifying on several bills, including LD 1101, "An Act to Address the Limited Availability of Counsel for Indigent Parties." The bill seeks to address the Sixth Amendment crisis brought to light in our Robbins case by increasing funding for indigent defense services. The Sixth Amendment requires the state to provide counsel to people charged with a crime who cannot afford their own, yet hundreds of people are actively being denied that right in Maine.

 

We shared with lawmakers that setting up new public defender offices throughout Maine will help ease the crisis by increasing the number of qualified and effective defense attorneys in the state. This approach offers a more reliable alternative to the state's current system, which largely depends on contracting private defense attorneys who are often unavailable and, at times, unable to adequately represent their clients.

Read the full testimony here

 

Relevant Reads

• Bangor Daily News: What comes next after ruling determines Maine is violating people’s Constitutional rights
• Portland Press Herald: Maine has spent millions to digitize court records — and they’re still mostly not online
• Maine Morning Star: How Trump carved a pathway for his mass deportations through executive orders
• Maine Public: Emergency bill to shore up public defender system lacks support from Maine civil rights advocates