From the institutes, labs and civil-society groups.

A detailed record of federal courts and grand juries rejecting the Trump administration’s use of subpoenas, search warrants, and indictments. The post “Zero Evidence”: How Judges and Grand Juries Have Rejected the Trump Administration Efforts to Investigate and Prosecute appeared first on Just Security .

These EU AI Act provisions will start to apply on August 2 | Edition #307

The singularity will be seen in hindsight as an interregnum

Data centers have become a kind of boogeyman among voters lately. Recent polling even suggests that new data center construction is less popular than the IRS. Only 27% of Americans at least somewhat ...

Signup to receive the Early Edition in your inbox here. A curated guide to major news and developments over the weekend. Here’s today’s news: IRAN WAR The Pentagon said yesterday that a third member of the U.S. military had died over the weekend during the disposal of a downed Iranian attack drone in northern Iraq […] The post Early Edition: July 20, 2026 appeared first on Just Security .

Rapidly growing agetech industry has a significant opportunity to close trust gap, increase product adoption with increased transparency WASHINGTON, D.C. — (July 20, 2026) — The Future of Privacy Forum (FPF) — a global non-profit focused on data protection, AI, and emerging technologies — today released findings from a comprehensive March 2026 survey about how […]

Greetings from a world where…

A recent story of a writer who was mistakenly identified, tracked, and arrested using data from Flock cameras has gone viral. The New Jersey plates that were allegedly stolen from the LA dealer were 34 03 DTM , not 34 10 DTM . But when the police report was created and the plate was entered into Flock’s system, it was just recorded as 34 DTM . Just the five large characters, no little number in the middle. And Flock’s AI tech wasn’t registering that non-standard little number when it began picki

Erin Davis calls it the “SuperDuperPOD.” That’s two things in one name: pharmaceutical giant Bristol Myers Squibb (BMS) already runs one of the largest AI clusters in life sciences, with serious results to show for it. And they’re doubling down. BMS announced today it is deploying its second NVIDIA DGX SuperPOD, this one built on […]

OpenAI shares lessons from deploying long-running AI models, highlighting new safety risks, observed failures, and improved safeguards through iterative deployment.

WASHINGTON – The American Civil Liberties Union today kicked off its largest ever National Advocacy Institute, hosting over 900 students from all 50 states, Washington, D.C., and Puerto Rico in the ...

On Wednesday 15 July 2026, the Court of Appeal (Vos MR, Warby LJ and Whipple LJ) handed down judgement in the appeal of Vince v ANL [2026] EWCA Civ 899. Eco entrepreneur, Dale Vince brought a defamation and subsequent data protection claim against Associated Newspapers in relation to the publication of an article which juxtaposed […]

4th IEEE International Conference on Medical Artificial Intelligence [Zhengzhou, China] [Nov 20, 2026 - Nov 22, 2026]

Long video summarization presents significant challenges for multimodal large language models (MLLMs), particularly in maintaining temporal fidelity over extended durations and producing summaries that are both semantically and temporally grounded. We introduce LVSum, a human-annotated benchmark for evaluating long-form video summarization with fine-grained temporal alignment. LVSum comprises 72 diverse videos spanning 13 domains with an average duration of 16 minutes, each annotated with up to

Want proof that policymakers and school staff need help with artificial intelligence? Look no further than the Los Angeles Unified School District. Superintendent Alberto Carvalho resigned June 21 as ...

PsychAdapter lets researchers dial in on personality traits, age, and mental health characteristics to generate text that sounds like real individuals, opening the door to training simulations and personalized content.

From designing new antibodies to simulating 1,000 years of climate in a day, AI is transforming what's possible—but humans remain the ones deciding what matters.

Leveraging statistical concepts from measurement science and education, AI researchers have greatly reduced the computational demand of predicting how the largest of large language models will scale up in the future. It could save millions of dollars in training costs.

Plus: Palantir’s power, cancer vaccines and AI on iPhones

Lots of articles about this . As usual, you can also use this squid post to talk about the security stories in the news that I haven’t covered. Blog moderation policy.

Senior research fellow Gary Miller spoke to Financial Times about attempts to exploit mobile network vulnerabilities to track US personnel during the Iran war. The post US Military Smartphones Targeted Through Roaming and Ad Tech appeared first on The Citizen Lab .

Reversing course, Flock Safety—the surveillance technology vendor most known for its extensive network of automated license plate readers — has announced that it will end a pilot for its acoustic gunshot detection devices to identify signs of “human distress.” In October 2025, EFF warned the public that Flock was rolling out a new feature called “Distress Detection” that would be deployed through their acoustic gunshot detection devices (formerly known as Flock Raven, now called Audio Detection)

I watched Ferris Bueller’s Day Off which was as great as I remember. It has an age rating of PG-13 in the US and 15 in the UK for bad language and some kissing. (There is also lying to parents but apparently that doesn’t count.) This seems absurd. I have way less problem with some light cursing/petting than films that are ostensibly for kids like cartoons and superhero movies that treat mass existential threat as an incident plot point in the hero’s journey. Like: Frozen 2 in which the entire ci

In some countries, the debate around technological sovereignty has been framed as a race toward independence. Governments have invested billions in sovereign cloud initiatives, national AI programs, semiconductor manufacturing, and critical infrastructure under the implicit assumption that sovereignty means self-sufficiency. But that assumption is detached from reality.

When Ubisoft's Watch Dogs 2 was released in 2016, it was a headtrip for those of us working on digital-rights issues in the Bay Area. During the day, I'd fight tech-authoritarianism from EFF's San Francisco offices and then, at night, I'd fight tech-authoritarianism in an uncanny simulation of San Francisco from my home gaming console. Watch Dogs 2 is an open-world video game that follows a hacktivist collective called Dedsec as they take on surveillance tech and discriminatory AI systems that a

Lowest cost per token from extreme codesign maximizes intelligence per dollar for post-training in the agentic era.

Google introduces Gemini 3.5 Flash Cyber, a lightweight cybersecurity model to find and patch vulnerabilities.

Brazil may soon become the first country in the Americas to regulate digital platforms before they do anything wrong. Bill 4,675/2025 would let the Administrative Council for Economic Defense (CADE) designate large technology companies as having “systemic relevance” and impose special obligations without first proving anticompetitive conduct or consumer harm. That is a substantial change ... Brazil’s Gatekeeper Bill Gets a Tuneup, Not a Rethink The post Brazil’s Gatekeeper Bill Gets a Tuneup, No

The biologist’s bold “energetic view of life” looks to the body’s strangest organelles as the link between cells, health, and mind and the foundation of our experience of being alive. The post Martin Picard’s Mitochondrial Theory of Mind first appeared on Quanta Magazine

As usual, part 2 of the weekly deals with speculative, regulatory, political and alignment questions.

The standard AI competition story has plenty of supposed villains. It just keeps leaving out one of the biggest. Regulators and academics warn that a small group of firms—including Amazon, Anthropic, Google, and OpenAI, with Microsoft and Meta sometimes added—will dominate generative artificial intelligence. Their advantages in computing power, capital, data, and distribution will harden ... The Missing Rival: China and the Limits of AI Antitrust The post The Missing Rival: China and the Limits

Artificial Intelligence can produce biased outputs. In part, this is because unrepresentative data is used to train, validate and test AI. To remedy skewed datasets and train fairer AI, many call for more comprehensive and systematic data production and processing about diverse people’s bodies and lives, including disabled people. Yet this response rests on a number of assumptions. Drawing on disability data, I argue that we should be cautious about these assumptions when regulating AI. The post

On 15 July 2026, the Court of Appeal handed down an important judgment in Vince v Associated Newspapers ([2026] EWCA Civ 899). This decision establishes, for the first time, that a person can sue a newspaper under data protection law for an “unfair” article even if, under the law of defamation, there is no claim. […]

The post Privacy and human rights award 2026 appeared first on Access Now .

CEPS is pleased to announce a call for interest for a new three-day executive course designed for mid- to senior level health policy experts and practitioners from low- and middle-income countries ...

The post Why we’re increasing the AI focus of our job board appeared first on 80,000 Hours .

An intelligence official on what the declassified material shows—and does not show—about foreign threats and the president’s allegation of an intelligence community cover-up. The post An Initial Readout of Trump’s Election-Interference Speech appeared first on Just Security .

Suggestion for hyperscalers feeling pressure over data center water use: Buy up a few exclusive country clubs, convert the golf courses into public parks, pay for guides and binoculars to get the previous members into birdwatching - help them embrace a more sustainable hobby! Google used 10.9 billion gallons in 2025 , so about 30 million gallons per day. The Coachella Valley has 120 golf courses each using ~800 acre-feet per year , which is ~750,000 gallons per day. So Google buying up 40 of tho

As concerns around data privacy in machine learning grow, the ability to unlearn—or remove—specific data points from trained models becomes increasingly important. While state-of-the-art unlearning methods have emerged in response, they typically treat all points in the forget set equally. In this work, we challenge this approach by asking: do points that have a negligible impact on the model’s learning need to be removed? Through a comparative analysis of influence functions across language and