Securus Jail Call Monitoring, Cities Lose Control Over Surveillance, Police IDs Made from Video, Nina Loshkajian Answers 5 Questions & More

Vol. 7, Issue 6

The Digital Forensics Unit, Gregory Herrera, Shane Ferro, and Joel Schmidt

Jun 01, 2026

June 1, 2026

Welcome to Decrypting a Defense, the monthly newsletter of The Legal Aid Society’s Digital Forensics Unit. This month, Gregory Herrera reports on New York City’s intent to continue contracting with Securus to surveil incarcerated New Yorkers. Then, Shane Ferro discusses how cities lack control over the surveillance tools they contract with private companies to use. Next, Joel Schmidt reviews recent court decisions on the sufficiency of a police officer’s familiarity with an individual to identify them in a video. Finally, our guest Nina Loshkajian answers five of our questions.

Job Postings

The Legal Aid Society’s Digital Forensics Unit has two openings, one Staff Attorney position and one Analyst position. For over a decade, we have used technology to advocate for our clients in courtrooms across New York City and have fought against government surveillance and the erosion of digital privacy rights. As the the use of electronic evidence has grown, so has the need for our work. If you are interested in working with us on digital forensics and electronic surveillance related issues, and you meet the required qualifications, please apply.

Staff Attorney, Digital Forensics Unit (applications close at 3:00pm EST on June 8, 2026)
Digital Forensics Analyst, Digital Forensics Unit (applications close at 3:00pm EST on June 5, 2026)

In the News

A photo of three rotary pay phones on a wall. — Three phones by Pavan Trikutam via Unsplash

Keeping Families Together, One Data Point at a Time

Gregory Herrera, Digital Forensics Staff Attorney

On May 8, 2026, the Mamdani Administration published its intent to extend its current contract with Securus Technologies for another 6 months while a new contract is negotiated. Securus is one of the two biggest prison telecommunications companies, providing the infrastructure for incarcerated individuals to communicate with their loved ones via phone calls, video calls, and messaging. Among their other offerings are money transfer services, tablets for incarcerated people to access educational content, and services to help prisons deal with “the growing threat of contraband mobile phones.”

Their initial five-year contract with New York City to provide their services in Rikers Island jails started in 2014 and has been extended several times since. Back in mid-April of this year, the City gave notice of the new contract it’s negotiating with Securus. Though it’s not finalized, it’s currently envisioned as a five-year contract starting in July 2026 (with another five-year renewal clause) and costing $23.2 million.

Many privacy supporters, public defenders, and civil rights advocates are urging Mayor Mamdani to dump Securus, citing its past controversies and potential abuse of the vast quantities of personal data in its possession, as documented in a recently released report from the Surveillance Technology Oversight Project. Every call someone incarcerated at Rikers makes to loved ones is monitored in real time, recorded, and transcribed into searchable text by AI-powered cloud infrastructure. On top of that, their Investigator Pro tool captures callers’ distinct vocal and speech patterns to create “voiceprint” profiles. All this data is analyzed in THREADS and combined with money transfer and tablet data to map out social networks. According to a recent public comment [PDF] submitted by Brooklyn Defender Services, this means more than one million calls analyzed per day. This massive trove of data is fed into Securus’s artificial intelligence (AI) algorithm.

Advocates are also concerned that this data will be shared with federal Immigration and Customs Enforcement (ICE), as Securus’s THREADS can integrate with Palantir’s tools. Palantir has custom-built tools for ICE like ImmigrationOS, and all the data Securus has collected from incarcerated people at Rikers can put immigration communities at risk. If you thought that a simple fix would be contractual language to force Securus to protect certain data or follow certain laws, you’d be wrong. Phone calls between incarcerated people and their lawyers – a specially protected category of communication – were repeatedly recorded and shared with prosecutors and the police.

For over a decade, Securus has surveilled incarcerated New Yorkers and their family members who have received their calls, and now it is using every scrap of data it can get to feed its AI models. The Mamdani administration has the chance to chart a new path.

A photo of two surveillance cameras on the side of a building with a blue-sky background — Photo by Arthur Mazi on Unsplash

Democracy Dies in Contracts

Shane Ferro, Digital Forensics Staff Attorney

Most surveillance technology is not owned by the cities that use it, and its use is controlled by third-party contracts negotiated years into the future. This week, 404 Media reported that the city of Dayton, Ohio is trying to get rid of its Flock automated license plate reader cameras (after residents found out that the city was sharing license plate data with immigration enforcement “apparently on accident”), but isn’t sure how. The contract that the city has with Flock apparently does not allow the city to simply take down the cameras. Instead, the “Dayton Police Department agreed to work with Public Works to put bags over the cameras” in an attempt to stop the cameras from capturing license plate data that neither the city nor its citizens want recorded, according to 404.

When a similar thing played out in Evanston, Illinois in 2025, the city took down its Flock cameras only to have the private company put them back up without the city’s permission. According to the Evanston Roundtable, the city of Evanston had approved a 5-year contract extension with Flock in 2024, and when the city tried to terminate the contract a year later (with nearly $150,000 owed to Flock for the remainder of the contract) the company fought back. Although litigation has yet to be filed, at least two Flock cameras were still up in the city of Evanston until the end of March 2026. They weren’t taken down until a reporter started making calls asking about them.

Meanwhile, in Chicago, the Cook County Sheriff is pushing the county to sign new contracts for both Flock cameras and a dystopian AI system to be put into the county jails, without having gone through the “county’s standard public notice and competitive bidding requirements,” according to the Chicago Reader. AI systems are now obviously necessary on top of the data-collecting tools, because so much data has been collected that AI is required to make sense of any of it.

BriefCam, the AI system the sheriff wants to put in the jails, claims to hook up to existing surveillance systems, then identify and quantify every face, object, and license plate that exists on the video footage. Its marketing brochure shows a busy crosswalk, with cars, motorcycles, and pedestrians crossing. All three categories, including people with their backs to the camera, are seemingly identified and confirmed to be “registered.”

Registered to whom exactly?

In the Courts

A photo of the U.S. Supreme Court building. — Photo by Claire Anderson on Unsplash

Can a Non-Eyewitness Police Officer Identify a Person on Video as the Defendant? The Appellate Division Establishes Limits

Joel Schmidt, Digital Forensics Staff Attorney

In 2024 we reported on the Court of Appeals decision People v. Mosley in which the New York high court held that before a non-eyewitness may testify that a person depicted in surveillance footage is the defendant the judge must first be satisfied that the witness is sufficiently familiar with the defendant to be able to identify them and that the jury would benefit from such testimony. When considering the prior familiarity prong the high court issued guidance for trial judges to consider whether the witness is generally familiar with the defendant’s appearance, for how long they’ve been familiar, in what circumstances, whether the witness was familiar with the witness at the time the footage was recorded, whether the witness was familiar with how the defendant customarily dressed, and whether the person depicted in the surveillance footage exhibits certain traits unique to the defendant such as a distinctive gait, scar, or tattoo.

Earlier this year we reported on the Appellate Division, First Department decision People v. Martin in which the intermediate appellate court held a trial court violated Mosley’s proscriptions when it allowed a police officer who was not a witness to the crime to identify a person on video as the defendant because “[t]he officer did not know defendant before defendant’s arrest and spent time with him only while defendant was in custody.” My colleague Shane Ferro explained Martin’s significant ramifications on the typical facial recognition case in New York City, where even the NYPD is compelled [PDF] to admit that a facial recognition match cannot constitute probable cause to make an arrest (for good reason), and instead frequently relies on an identification from a non-eyewitness police officer who previously arrested or interacted with the defendant.

Martin built on People v. Dean, wherein the Appellate Division, Second Department held a trial court “improvidently exercised its discretion in permitting the sergeant’s testimony for the purpose of identification of the defendant as the individual depicted in stills from surveillance footage from the dry cleaners” because “the sergeant’s familiarity with the defendant was so limited that there was no basis for the court to conclude that the testimony would be of assistance to the jury in determining if the defendant was the individual depicted in the surveillance video.”

We now bring you news of People v. Jones in which the Appellate Division, First Department found no Mosley violation, holding “The court did not abuse its discretion in permitting a police detective to give lay opinion testimony identifying defendant and his brother in several segments of surveillance footage showing them in the lobby of their building immediately before and after the shooting” because “[t]he detective had known defendant and codefendant for many years based on his long tenure in the 49th Precinct, which encompassed the Pelham Parkway Houses, where the defendants resided.”

Martin and Dean contrast and compare well with Jones. Whereas Martin and Dean illustrate what does not suffice to meet the Mosley prior familiarity prong, Jones illustrates the kind of extensive prior familiarity that would justify such testimony (“The detective had known defendant and codefendant for many years.”) Anything in between is arguably not the kind of prior familiarity contemplated by the high court in Mosley. Taken together, Martin, Dean, and Jones serve as an important Appellate Division message to trial courts: Don’t water down Mosley.

5 Questions with Nina Loshkajian

5 Questions is our take on the 21 Questions game. We’ve sent twenty-one questions to specialists in digital forensics, surveillance, and technology and requested that they answer any five. This month’s participant is Nina Loshkajian.

Nina Loshkajian (she/her) is the Technology & Racial Justice Collaborative Fellow at the Center on Race, Inequality, and the Law at NYU Law, where she leads the Center’s work at the intersection of race and technology. Prior to joining the Center, Nina worked as a Staff Attorney for the Surveillance Technology Oversight Project (S.T.O.P.).

1. What was your career path to get where you are now?

I went to law school with the goal of pursuing a career as a civil rights lawyer. During my time at NYU Law, I had the opportunity to explore this work from a number of different angles through my clinics and internships. During my 1L summer, I interned with the Brennan Center’s Liberty & National Security Program. While there, I engaged in advocacy in support of the POST Act in NYC. This experience was pivotal in opening my eyes to the harms of surveillance and discriminatory technology, which is what I’ve ended up spending my career focused on. After law school, I spent three years at S.T.O.P. engaged in legislative and legal advocacy to combat systems of mass surveillance, and in November 2024 I returned to NYU Law to work with the Center, where I tackle these issues with a specific focus on racial justice implications, addressing the disproportionately high levels of surveillance borne by communities of color.

2. What do you do to keep up with recent developments in digital forensics and electronic surveillance issues?

There are many great newsletters that help me keep my finger on the pulse. In addition to Decrypting a Defense, which is honestly one of my favorites, I recommend the Electronic Frontier Foundation’s EFFector and the Center’s own Technology & Racial Justice Collaborative quarterly newsletter. I also can’t say enough good things about 404Media — I have learned so much from their investigative reporting.

3. How does your knowledge of electronic surveillance impact your day-to-day choices about privacy?

While I have definitely implemented a lot of important personal digital hygiene practices from all I’ve learned doing this work (regular software updates, using Signal, etc.), I also firmly believe that we need more protections from the top down and I try not to guilt others or myself too harshly for not being perfect all the time. I liken this to how individuals sometimes are guilted for not always recycling or composting, when in reality corporations are the top polluters worsening climate change and there are insufficient structural mechanisms in place to support individuals wanting to make better choices. In the fight for privacy, I think it is important to do what is within our control, but we should recognize where personal control over our data has been corroded and thus also channel energy towards demanding comprehensive data privacy laws and protection from predatory corporate practices.

4. What are some of your favorite software and tools that are free of charge? How do you use them?

For anyone trying to stop using the same one password from memory for all your accounts (which I definitely recommend!), there are multiple websites offering free diceware that can help you generate unique passwords for different websites. I use this along with a password manager and it definitely gives me a little more peace of mind.

5. What is the most rewarding part of your work? What is the most frustrating part of your work?

By far the most rewarding part of my work is the collaborative nature of it — getting to be a part of a wonderful community of people who care deeply about protecting privacy and reining in technological harms is an incredible privilege. This community helps me stay sane through the more frustrating parts of the job, from countering false narratives touted by proponents of surveillance technology to dealing with how slowly legislative bodies often move due to political dynamics out of our control. It is motivating to work in coalition with other organizations and individuals who inspire me daily with their passion and intelligence.