Ahead of a crucial hearing in Anthropic's legal standoff with the Pentagon, outside groups filed briefs largely rebuking the Trump administration's effort to restrict government use of the company's products. Here's a breakdown of what legal experts and advocates had to say.

Many thanks to @justinhendrix.bsky.social for hosting @jessicasilbey.bsky.social and me for the @techpolicypress.bsky.social podcast to talk about our paper "How AI Destroys Institutions." We really enjoyed the conversation.

Vietnam is Southeast Asia’s first country with a comprehensive AI law. Modeled on the EU’s risk-based approach, it emphasizes human oversight, risk classification, and digital sovereignty as part of Hanoi’s 2045 development push, reports Tech Policy Press fellow Lam Le.

For this week's Tech Policy Press podcast, I spoke to Woodrow Hartzog and Jessica Silbey about their forthcoming law review paper, "How AI Destroys Democratic Institutions." They say the "affordances of AI systems extinguish" key features of democratic institutions.

The ultimate significance of AI will not be measured in processing speed or productivity gains, but in whether it alters our relationship to time, writes R Street Institute president & co-founder Eli Lehrer. In the long arc of history, structural change announces itself when ordinary rhythms shift.

"the decline in [academic freedom in] the United States has been more rapid and pronounced, particularly in the last year. It has deteriorated by 50% from 2015 to 2025 and is now evaluated as moderate autonomy by country experts" academic-freedom-index.net/research/Aca...

"The differences between the fates of detained immigrant parents under the two presidents are stark, our analysis shows. The impact on mothers is particularly pronounced. Trump is deporting about four times as many moms of U.S. citizen children per day as Biden did."

‘Invasive’ AI-led mass surveillance in Africa violating freedoms, warn experts Countries across the continent have spent more than $2bn on Chinese tracking technology that is not ‘necessary or proportionate’, new report finds www.theguardian.com/global-devel...

I spoke with @rollingstone.com about the Trump administration's new AI framework. At a moment when a majority of Americans are asking for AI guardrails, this proposal moves in the opposite direction. Rather that constrain companies, it releases them from accountability. 🧵 www.msn.com/en-us/news/p...

For this week's Tech Policy Press podcast, I spoke to Boston University law professors Woodrow Hartzog and Jessica Silbey about their forthcoming law review paper, "How AI Destroys Democratic Institutions." They say the "affordances of AI systems extinguish" key features of democratic institutions.

"But Silicon Valley billionaires are not our sages; they’re our enablers, keeping us distracted and dumb, and making sure we never stop scrolling for long enough to think about why we are wasting our lives on their platforms."

I spoke to Marketplace "Make Me Smart" podcast host and senior Washington correspondent @kimberlyadams.bsky.social about the Anthropic-Pentagon dispute and what it tells us about where tech firms are drawing their "red lines" when it comes to military applications of AI.

I wrote a lot about Bob Mueller across a lot of chapters of his career. I wanted to share two @wired.com pieces that I wrote about his defining life experiences—both a picture of an era of public service all but lost today: 1) His brave *voluntary* service in Vietnam: www.wired.com/story/robert...

"A U.S. federal jury found Elon Musk liable on Friday for claims he defrauded Twitter shareholders by ‌trying to drive down the social media company's stock price so he could renegotiate or back out of a $44 billion takeover in 2022."

"Judge Paul Friedman, of the U.S. District Court for the District of Columbia, also ordered the Pentagon to restore the press passes of seven journalists for The Times."

The Trump administration unveiled a new National AI Legislative Framework today, outlining its approach to establishing a unified federal standard for artificial intelligence governance. Ben Lennett discusses the framework and a legislative proposal from Sen. Blackburn, the “TRUMP AMERICA AI Act.”

In an op-ed in the FT, French AI startup Mistral cofounder and CEO Arthur Mensch proposes "a revenue-based levy that would be applied to all commercial providers placing AI models on the market or putting them into service in Europe, reflecting their use of content publicly available online."

Google started rewriting headlines for videos in search results years ago; perhaps not a surprise it would do so now for news organizations, but still wonder what it means for a platform to re-author content in this way.

"The weeks-long investigation, details of which have not previously been reported, uncovered no records indicating censorship, according to two people familiar with the matter, who spoke on the condition of anonymity for fear of retribution."

2026 is poised to be an inflection point in Brazil for efforts to regulate AI in elections, but there are lingering questions about whether a key court will be able to keep pace with the issue, Andressa Michelotti and Tatiana Dourado write.

I spoke to Marketplace "Make Me Smart" podcast host and senior Washington correspondent @kimberlyadams.bsky.social about the Anthropic-Pentagon dispute and what it tells us about where tech firms are drawing their "red lines" when it comes to military applications of AI.

I wrote about why @knightfoundation.org invested in @bsky.app: philanthropy can act when mission & opportunity align. We believe in @jay.bsky.team's vision for a better social web—& in ideas (like @masnick.com's) that grow out of the research ecosystem Knight has long supported. kf.org/kfbluesky

“At the end of the day, if you're outside the US and you're a government, the bottom line is very clear. You have to control your own technology. Otherwise, you're going to be dependent on both the US government and US tech.” www.techpolicy.press/anthropic-pe...

Hegseth said he was cutting Anthropic off from the military and its partners "effective immediately" and that the transition to its competitors would be "seamless." Insiders tell @reuters.com a different story. Here's our story, with Mike Stone and Alexandra Alper: www.reuters.com/business/heg...

To give you a sense of what’s actually going on in the eastern Mediterranean front in this war, right now people in Beirut are basically hoping that an algorithm doesn’t decide someone in their building might be connected to Hezbollah.

Spyware intermediaries facilitate global spyware deals that would otherwise be blocked under regional export rules. A new Atlantic Council report explores their role, reports Tech Policy Press fellow Vas Panagiotopoulos.

The combination of the Anthropic-Pentagon clash and the US and Israel’s joint strikes on Iran have turned a years-long theoretical debate about military applications of AI into an urgent one, writes Tech Policy Press fellow James Ball. That debate is being had in capitals far from Washington.

In high-risk communities, online conflict accelerates offline retaliation, write Thomas Abt, Desmond Patton, and Joseph Richardson. Prevention requires coordination and commitment from tech platforms, community violence interrupters, researchers, law enforcement and impacted youth.

It's been two months since the Senate passed the DEFIANCE Act, which would allow victims to sue people who make nonconsensual deepfake AI porn of them, by unanimous consent but the House has not even taken it up for a vote. AOC, Josh Hawley and Lindsey Graham support it. But so far, nothing.

A deal to integrate OpenAI's products into classrooms in California’s biggest public university system has sparked backlash from students and faculty, presaging similar battles over AI in higher education around the country, Tech Policy Press fellow Chris Mills Rodrigo reports.

this is interesting reporting but i think it’s worth pointing out that even if ai were to blame, humans would still be to blame www.semafor.com/article/03/1...

A new report from the Atlantic Council shows brokers and resellers facilitate global spyware deals that would be blocked under regional export rules, reports Tech Policy Press fellow Vas Panagiotopoulos. www.techpolicy.press/policymakers...

"For 19 days, the majority of Iran’s 92 million people have been cut off from the outside world, according to watchdog groups that track internet censorship in the country, as the country’s regime has tried to suppress communication and maintain control over its population during wartime."

In high-risk communities, online conflict accelerates offline retaliation, write Thomas Abt, Desmond Patton, and Joseph Richardson. Prevention requires coordination and commitment from tech platforms, community violence interrupters, researchers, law enforcement and impacted youth.

"Silicon Valley made risky bets in recent years on developing defense-related technology and providing services to the U.S. military establishment. Now those bets are paying off."

A wave of AI-generated images and video has emerged from the Iran war. But a new phenomenon is also observable, write shirin anlen and Mahsa Alimardani from WITNESS: the spread of technical-looking analyses used to discredit authentic evidence.

1/ I've taught/referenced three democracy indexes/reports over the years- from Freedom House, the V-Dem Institute, and The Economist Intelligence Unit. V-Dem just came out with its 2026 report. Like most measures of the health of democracy, it's not good news. www.v-dem.net/documents/75...

Japan’s new political party, Team Mirai, illustrates the viability of a different way to do politics, write Nathan Sanders and Bruce Schneier—one in which technology is used to make democratic processes stronger, instead of undermining them.

"The Department of Homeland Security is poised to pour funds into surveillance technology projects this year amid signs of lagging governance."

1/ I've taught/referenced three democracy indexes/reports over the years- from Freedom House, the V-Dem Institute, and The Economist Intelligence Unit. V-Dem just came out with its 2026 report. Like most measures of the health of democracy, it's not good news. www.v-dem.net/documents/75...

shirin anlen and I at @witnessorg.bsky.social did a deep dive into viral discrediting of authentic documentation of civilian casualties in Iran as AI-generated. The first case, fake heatmap analyses of real Tehran strike photos, spread widely. Many people I know believed it. That's the problem.

"Several school districts are already using AI cameras, including the state's largest district. Newark schools signed a $12 million contract to install 7,700 cameras to watch its 44,000 students. The AI cameras were up and running at the start of this school year." nj1015.com/new-jersey-s...

The AI bubble is a simple story, like the housing bubble, even if the NYT and other elite media outlets want to tell you otherwise cepr.net/publications...

A wave of AI-generated images and video has emerged from the Iran war. But a new phenomenon is also observable, write shirin anlen and Mahsa Alimardani from WITNESS: the spread of technical-looking analyses used to discredit authentic evidence.

Recommended read from @courtneyr.bsky.social on the Coalition for Independent Technology Research lawsuit against the Trump administration for chilling speech and retaliating against researchers working in the public interest to study phenomena on tech platforms. (I am a member of the Coalition.)

Long read on DOGE and Musk from @bentarnoff.com and @quinnslobodian.com this morning. "Muskism proved not a governing philosophy but a toolkit available to those who govern," they write.

Here's @techpolicypress.bsky.social with details on the forthcoming Presidential Decree on Brazil's new child protection law. Huge news that is mostly going unnoticed in US media. www.techpolicy.press/brazil-wants...

👀

Brazil will soon find out if its new law aimed protecting children online will stick. A copy of a draft enforcement decree obtained by Tech Policy Press offers a fuller picture of how it may be implemented, fellow Tatiana Dias reports.

"When a mother from eastern Tennessee asked local police how someone had created naked photos of her teenage daughter, she recalls being told it was a company she’d never heard of: xAI, the artificial intelligence start-up run by Tesla CEO Elon Musk." www.washingtonpost.com/technology/2...