Nathalie Maréchal asks, why are we betting the economy and the future of democracy on cult-like beliefs?

The US Federal Government Inventory Still Needs an Accessible Interface

What if AI Agents Could Hack the Legal Forums Meant to Restrain Them?

Call for Proposals to Host Computation + Journalism (C+J) Symposium 2027 Since 2008 the Computation + Journalism (C+J) Symposium has brought together a unique interdisciplinary community of journalis...

Thomson, the company's proprietary LLM, was trained and is run at a fraction of the cost of comparable frontier models and remains fully owned and controlled by Thomson Reuters

In a random sample of 10,000 webpages collected in July 2026, one-in-ten show signs of being written or substantially edited by AI.

Frontier large language models (LLMs) have rapidly developed from helpful coding assistants to highly capable cybersecurity systems. After several security incidents in the past few months, more overs...

San Francisco, CA

The US Federal Government Inventory Still Needs an Accessible Interface

📍map

The growing number of AI agents roaming the internet will eventually force us to verify what the old web mostly presumed: that there is a morally and legally accountable person somewhere in the chain.

The idea that artificial intelligence can “reason” is more intuitive than ever. But intuitions can be wrong, and the science is far from settled.

On Monday, the Society of Professional Journalists (SPJ) dropped new proposed revisions to its code of ethics — the first revision to one of the most-cited media ethics standards in the country since ...

The A.I. revolution is stalling because companies don’t want to admit that integrating the technology is expensive and slow and requires human effort.

Governor Hochul and Attorney General James announced the release of final rules on how social media companies must restrict addictive features on their platforms to comply with the Stop Addictive Feed...

New research on guidelines for evaluating AI guidelines

You don’t need Claude Science to support literature reviews

The Department of Communication at Stanford University seeks applicants for a junior appointment at the rank of tenure-track Assistant Professor. We invite applications from individuals working at the...

The AI Accountability Review translates research to help bridge the gap between knowledge and practice for AI policymakers, practitioners, and researchers. Click to read AI Accountability Review, by N...

As the rapid proliferation of AI systems and harms spurs efforts in AI governance around the world, prioritizing among competing policy options has become increasingly challenging for policymakers and...

The unit’s future was thrown into doubt after its public work was halted despite winning praise from AI developers.

A German regional court has ruled that Google is directly liable for the content of its AI search overviews. According to the court, previous limited liability protections for search engine operators don't apply to AI overviews. In this case, Google's AI had falsely linked two publishers to fraud and made claims that didn't appear in any of the linked sources. The ruling could set a precedent for AI-generated content liability worldwide.

Features for humans using AI agents and the AI agents themselves, including "ask" AI buttons and an MCP server

Our progress toward recursive self-improvement, and its implications.

To move from high-level values to effective accountability, we still need to bridge the gap between abstractions and quantifiable, data-driven metrics.

By the authority vested in me as President by the Constitution and the laws of the United States of America, it is hereby ordered: Section 1.  Purpose.

We’re hiring our next Head of AI. This is a rare and compelling opportunity, at a moment when our information environment is under significant strain.

An Open Competition for Investigative TeamsThe Agentic AI Investigative Journalism ChallengeCoding agents can now work through messy document dumps, sur...

Deal marks first significant delay of digital rules amid pressure from the U.S.

How goblin outputs spread in AI models: timeline, root cause, and fixes behind personality-driven quirks in GPT-5 behavior.

AI and Ethics - Romance fraud has undergone a structural transformation. What was once a labor-intensive confidence scheme, limited by the number of conversations a single operator could sustain,...

A decade of explainable AI research has produced important techniques for understanding AI models, but less clarity on who those explanations are for and what accountability goals they actually serve