Using an LLM with your secure messaging service, as Meta proposes, brings new risks to the architecture that protects privacy and security. www.aclu.org/news/privacy...

We can debate the moral complexities of individual responsibility for the misuse of LLMs, but policymakers need to be realistic that lapses are inevitable and take steps to limit certain uses of AI. www.aclu.org/news/privacy...

"Basically, he said that it’s a tool that isn’t going to be used — and if it is used, then 'you’re failing in the training department, big time.'" www.aclu.org/news/privacy...

@ACLU demanding more information on violations of photographer's rights, including DHS policy directives, training materials, legal analyses, and communications www.aclu.org/news/free-sp...

Perspective on AP report revealing how DHS uses AI and mass surveillance across the nation to target, detain, and seize money from innocent drivers www.aclu.org/news/privacy...

What happens when your device can only run what the government decides is OK? How we have gotten dangerously close to this kind of totalitarian control

A breakdown of the new “Mobile Fortify” face recognition program, including lists of a) The reasons it’s problematic, b) The vast array of data that the agencies can connect to c) Some of the unanswered questions that remain about the program.

Utah is the only state where political leaders are thinking through the potentially disastrous effects of the current effort to create a national digital ID through the driver's license system.

Exciting news: you can now subscribe to an ACLU newsletter to keep up with with ACLU technology policy blog posts and other news! Subscribe here: action.aclu.org/signup/free-...

A bureaucratic wall is being built between Americans and the information they have a right to access www.aclu.org/news/privacy...

We shouldn't get rid of police audit logs but we also shouldn't think they're worth more than they are

Some US prisons and jails have started using non-consensual, non-removable biometric wrist straps on prisoners to monitor their pulse, blood oxygenation level, skin temperature, activity level, and location. What does the ACLU think of this? I look at the issue in a post on our Free Future blog

Appalling... apparently "military grade" AI has fallen short...

News about the driver-surveillance company Flock just keeps coming in: findings from Massachusetts and Virginia, new deals with big companies, and more:

Digital driver’s licenses — now being built in many states — have a big problem that almost nobody is addressing: the threat of an explosion of ID demands from all quarters. There is a body of existing state laws that points toward a solution.

Generally people should have privacy from government, and the government should be an open book to the people — not the other way around.

Surveillance, AI, and airline ticket prices: hasbrouck.org/blog/archive... @jaystanley.bsky.social

There are a growing number of companies that are engaging in mass surveillance that could prove useful to the Trump Administration’s efforts to target immigrants. They have human rights responsibilities.

The coming battles over being identified versus not being identified

We are certainly not the ones whose power increases when companies crunch data with AI to individually set the prices they charge us.

A look at the prospects for, and reasons to support, open source LLM models and research www.aclu.org/news/privacy...

It's not hard to imagine police departments deploying fleets of self-docking, self-navigating patrol drones over communities. We're starting to see that in the commercial space now.

A look at the possible future of local AI models, and what that might mean for the question of how much of a centralizing force the technology will be. www.aclu.org/news/privacy...

Policymakers are beginning to recognize that you can't count on surveilling your local residents while keeping that data out of the hands of ICE and the Trump Administration www.aclu.org/news/privacy...

What are the chances that the ability to train a frontier LLM will become more democratized? That's a key question in evaluating whether this tech will become yet another source of tech leverage over ordinary people. www.aclu.org/news/racial-...

There's a lot going on with Flock and license plate recognition; here's an overview of how this authoritarian technology is expanding. Latest post on @ACLU Freefuture blog www.aclu.org/news/privacy...

Large language models have been evolving in dramatic — and, for those who care about freedom — suspenseful ways.

My first post on my new newsletter: If your local police start using Flock, they could target you just because some algorithm has decided your movement patterns suggest criminality, under new AI capabilities that the company has created. ourfreefuture.substack.com/p/surveillan...

Dear all, today I'm announcing "Free Future" a newsletter/blog that I'm launching on Substack, where I'll be publishing my writing for the ACLU on privacy and technology policy topics. I hope you will find it of interest! ourfreefuture.substack.com/about

Noteworthy endorsement from the American Enterprise Institute's Jim Harper of the ACLU's efforts to require that smartphone digital-ID systems operate on a decentralized basis--no location tracking, no phoning home, no kill-switch functionality--and support authorization short of identification.

I think Harper is correct in endorsing the work of folks like @jaystanley.bsky.social. He's also literally walked the walk on these issues, having accepted and completed a challenge in 2006 to fly from SFO to D.C. without showing any ID to airport screeners.

Press release from @jaystanley.bsky.social here: www.aclu.org/press-releas...

Not a great moment in history to let our law enforcement agencies adopt new authoritarian tracking technologies. www.technologyreview.com/2025/05/12/1...

Trump administration re-directing the nation's spy satellites from foreign countries onto US border communities. www.reuters.com/world/us/tru...

Excellent piece on the danger we face of getting locked into a decisively bad new ID tracking infrastructure: www.projectglitch.xyz/p/how-the-us...

Large language model technology is super-charging video analytics. Unless we enact guardrails, we can expect that nearly every rule, regulation, law, and employer dictate that can be enforced through the visual monitoring of human beings will be subject to unblinking AI cameras ready to report us.

A top NYPD official wants to use drones for routine patrol. Americans need to vehemently oppose that use of drones. No other PD is currently doing that (with possible exception of Beverly Hills). This after the dept has already used drones abusively to surveil protests and people's back yards.

We partnered with U.S. PIRG Education Fund in a new report mapping every state's status on comprehensive #privacy laws. ⚖️ Find out exactly what rights your state gives you — and what it doesn't. 🧭📄 epic.org/state-of-pri...

One of our periodic reminders that in addition to being bad for privacy and for low-income and marginalized people, a cashless society is brittle and susceptible to centralized points of failure (and control).

We're very skeptical about "Content Authenticity" proposals to lock down media hardware and software so we can prove provenance. They are unlikely to work, and if they did would be likely to entrench big tech players at the expense of software freedom.

A convincing argument that retail industry efforts to portray shoplifting as some great new problem are baloney -- that shoplifting rates have not changed in decades and that a deceptive new report was sponsored by company that sells video analytics and other security measures for stores.

The FAA has banned drones across wide swaths of New Jersey in response to the recent hysteria. This seems stupid.

The companies providing face recognition-based access control in sports stadiums want to push FaceRec from sports stadiums outward -- think, anywhere that you need a ticket, pass, or ID to enter. This might be good for assorted team owners, companies, and billionaires, but it's not good for us.

We’ve just published a 6-page white paper on why the use of AI to help draft police reports is a bad idea. We see four general problems with it: 1. AI problems 2. Evidentary and memory problems 3. Transparency and discovery problems 4. A loss of police reports’ human disciplinary function

Very nice piece on some of the problems with society going cashless, with focus on seniors. Not mentioned: cashless transactions can also be blocked by censorious middlemen.

Digital payments under our current system not only represent rent-seeking exploitation by oligopolistic companies per this article, they are terrible for privacy and exclude vulnerable people. Use cash and demand that places you shop accept it!

Excellent breakdown & super helpful for us tipping-point joiners

Maybe it's time that policymakers require that cellular location data be shielded from carriers using techniques like zero knowledge proofs (already demonstrated successfully) and from apps as well. www.aclu.org/news/privacy...