Compiler

Audrey Tang: A.I. should serve communities, not control them As Taiwan’s first digital minister, Tang pioneered a model of AI governance built on civic participation, transparency and “broad listening."

No one actually wants an all-seeing AI that disempowers humans, Aubrey Tang told @tekendra-parmar.bsky.social at #MozFest. Yet in Silicon Valley, both AI doomers & accelerators assume that future is inevitable. Audrey Tang has a word for this: hyperstition.

www.compiler.news/audrey-tang/

AI Labs Want More of Your Time. That's a Serious Problem.

In @compiler.news, CDT’s @mluria.bsky.social & Amy Winecoff warn that major AI labs are quietly optimizing chatbots for engagement — despite public claims to the contrary. As financial pressures rise, companies risk prioritizing stickiness over safety, even for vulnerable users.

A.I. labs want more of your time. That's a problem. Healthy disengagement — not stickiness — should become a core safety feature within AI chatbots.

If AI companies truly cared about “healthy engagement,” they’d design for healthy disengagement. Most aren’t, write @mluria.bsky.social and Amy Winecoff from @cdt.org.

Meta victory reveals cracks in antitrust enforcement If Meta’s not a monopoly, the ruling gives every other tech giant a green light to become one.

"If competition from TikTok and YouTube are enough to suggest that Meta’s products aren’t monopolistic, it says more about the limits of antitrust laws than it does Meta’s monopolistic nature," writes @tekendra-parmar.bsky.social.

UNESCO’s fight for relevance The United Nations cultural agency must prove multilateralism still matters in the age of AI.

Relevance was a key theme for #UNGA. For UNESCO it's even more pressing, writes @estherpaniagua.bsky.social. "The next director-general must ... preserve UNESCO’s relevance and navigate a world where faith in multilateral institutions is waning."
www.compiler.news/unescos-rele...

How to build AI to serve the Global Majority The lack of inclusion in global AI governance raises profound implications for socioeconomic growth and global power dynamics.

In "Building AI to serve the Global Majority", my latest commentary for @compiler.news, I discuss the impact of the lack of inclusion in global AI governance and development on socioeconomic growth and global power dynamics.

www.compiler.news/ai-for-globa...

Police disable safety features in AI crime tools Departments are turning off Axon’s A.I. safeguards in software that auto generates reports from body-cam footage, raising accountability concerns.

Great reporting from @tekendra-parmar.bsky.social reveals how police departments regularly deactivate safeguards meant to prevent bias and increase oversight of an AI tool designed to generate police reports.

www.compiler.news/axion-ai-bod...

We're co-hosting a special UNGA side event along with @humaneintelligence.bsky.social and meowtree.bsky.social of Technology Salon NY on 9/16 in NYC. Spend an evening with us digging into the big issues around how AI will affect the UN's SDGs.

Register to attend here: forms.gle/i2KM8roxJSUi...

Tech Policy Pop-Up: The UN Edition Join Compiler, Humane Intelligence & Technology Salon NYC for an evening of big ideas for building a safer, healthier global internet.

We're excited to announce the next edition of our global pop-up series on Sept. 16 in partnership with @humaneintelligence.bsky.social & Technology Salon NY. Details coming soon. For now, sign up for Compiler to get more info on this event, articles and more.
www.compiler.news/united-natio...

How to harness A.I. in the fight for democracy Autocratic regimes are using artificial intelligence to attack democracies, stirring concern from Romania to Rhode Island. But the technology could also become a powerful tool to safeguard elections if deployed in the right ways.

AI can threaten democracy—but it can also help defend it. Civil society groups are using AI to expand access to trustworthy info, fight deepfakes, and protect voters. Check out CDT @Isabellinzer.bsky.social analysis + recommendations in @Compiler.news: www.compiler.news/ai...

How to harness A.I. in the fight for democracy Autocratic regimes are using artificial intelligence to attack democracies, stirring concern from Romania to Rhode Island. But the technology could also become a powerful tool to safeguard elections i...

In a new op-ed for @compiler.news, CDT’s @isabelalinzer.bsky.social explores how AI can defend democracy from AI-fueled threats without undermining democratic principles – or missing out on its potential benefits. www.compiler.news/ai-democracy...

A great Friday read by my @cdt.org colleague @isabelalinzer.bsky.social laying out how global civil society groups leveraged AI to combat AI threats in 2024

How to harness A.I. in the fight for democracy Autocratic regimes are using artificial intelligence to attack democracies, stirring concern from Romania to Rhode Island. But the technology could also become a powerful tool to safeguard elections i...

Russia’s AI-powered interference in Romania’s election is a serious warning sign, writes @isabelalinzer.bsky.social of @cdt.org. The threat from AI is growing—but with the right tools, attention and resources, it can help defend democracy, not just endanger it.
www.compiler.news/ai-democracy...

I am thrilled share that I'm joining the @freeourfeeds.com team as interim director. Since leaving @globalvoices.org I've been focused on the incentives that influence how we talk to each other in public spaces and online. Free Our Feeds tackles this head on at the level of infrastructure.

What American really feel about A.I. The Trump administration's AI Action Plan received more than 10,000 comments. We analyzed all of them to find out how Americans feel about artificial intelligence and its impact on jobs, human rights,...

"While there are diverging views on how to regulate AI," wrote @tekendra-parmar.bsky.social, "it's clear that Americans are paying attention and more aware of how AI may benefit or harm their lives."

Read the full story here 👇

www.compiler.news/trump-ai-act...

The future of AI policy won’t be written by tech CEOs alone.

Thousands of Americans just made clear: They want innovation—but not without rules.

Ignore them at your peril.

Sentiment analysis revealed something surprising: Emotionally, Americans are split—pretty evenly—on AI.

There’s skepticism, yes. But also trust, curiosity, and a surprising amount of hope.

Anthropic warned of bioweapon risks.

Future of Life called for a moratorium on systems that could manipulate humans or evade control.

This was not a kumbaya moment among frontier labs.

Record industry groups (RIAA, SAG-AFTRA, etc) said: Scrap the race talk—just license the damn content.

EFF said even that’s not enough. Licensing deals often screw over independent artists and do little to stop biased or harmful AI.

Of course, companies had thoughts, too.

OpenAI used its comment to frame AI as a Cold War tech race with China, arguing for fewer regulations and more export controls.

Also, they’re getting sued by creators for copyright violations.

Many raised sharp questions about how the U.S. Copyright Office handles AI-assisted work.

They weren't just mad about stolen art. They were interrogating the rules, precedent and fairness.

Using machine learning and sentiment tools, we pulled themes from the comments.

Top words? “Data,” “work,” “rules” and “innovation.”

Top concern? Copyright—especially how creators’ work is being scraped to train models.

The public feedback revealed a deeply engaged, highly nuanced conversation.

90%+ of responses came from individuals, not companies.

Most weren’t doomsayers—they asked hard questions about labor, copyright and data.

🚨 More than 10,000 Americans weighed in on the Trump administration's AI Action Plan.

@tekendra-parmar.bsky.social analyzed all of it.

The results? Not a techlash—something more surprising.🧵

What American really feel about A.I. The Trump administration's AI Action Plan received more than 10,000 comments. We analyzed all of them to find out how Americans feel about artificial intelligence and its impact on jobs, human rights,...

We analyzed 10,066 comments to Trump's AI Action Plan to understand who sent feedback, what they said & how they felt. The results show an engaged public "aware of how AI may benefit or harm their lives," writes @tekendra-parmar.bsky.social.

www.compiler.news/trump-ai-act...

Join us: The Future of Online Expression In The Trump Era: Freedom vs Regulation" lunch with @ashkhen.bsky.social, Jenna Leventoff @aclu.org, Chris Marchese @netchoice.bsky.social, @lehogg.bsky.social
@fai.bsky.social.
📅 Wed, April 23
⏰ 12:00 PM
📍 Rayburn Building & Online
✉️ RSVP lu.ma/o51ptz6j

Compiler Tech policy news, analysis and opinion

And if you can subscribe to our newsletter and follow @compiler.news. As an editor, I'm always grateful to be edited. @elleryrb.bsky.social is always baller to work with. www.compiler.news

Kenyan lawsuit against Meta will test Big Tech accountability A Kenyan court will hear a $1.6 billion lawsuit that alleges Facebook helped incite genocide in Ethiopia. The case could be a model for those looking to hold platforms responsible for online hate and ...

The Kenyan case against Meta is one of the only measures of accountability the company faces for how its platform has been weaponized by military and paramilitary groups to destabilize nations and target ethnic minorities, writes @tekendra-parmar.bsky.social.

www.compiler.news/kenya-lawsui...

Kenyan lawsuit against Meta will test Big Tech accountability A Kenyan court will hear a $1.6 billion lawsuit that alleges Facebook helped incite genocide in Ethiopia. The case could be a model for those looking to hold platforms responsible for online hate and ...

@amnesty.org, @hrw.org & other groups accused the Ethiopian military of engaging in ethnic cleansing & war crimes against ethnic Tigrayans. A lawsuit alleging Meta played a role could set a global precedent by holding Big Tech accountable for real-world harms.

www.compiler.news/kenya-lawsui...

Kenyan lawsuit against Meta will test Big Tech accountability A Kenyan court will hear a $1.6 billion lawsuit that alleges Facebook helped incite genocide in Ethiopia. The case could be a model for those looking to hold platforms responsible for online hate and ...

The $1.6B suit against Meta brought by @marthadark.bsky.social & Foxglove Legal alleges Meta played a key role in fueling Ethiopia’s civil war by amplifying hate speech & failing to remove threats. If the case succeeds, it will set a global precedent.

www.compiler.news/kenya-lawsui...