AI Safety, Effective Altruism & Political Influence — Investigations Dossier

A guide to Horizon Accord’s investigations into the AI-safety movement, effective altruism, the institutions that fund it, and its growing political influence. Original reporting remains on its publication site.

Investigations and analysis

The AI Safety Network: Sex, Money, and the Politics of Extinction

The Berkeley rationalist community, funding, private networks, misconduct allegations, and AI-extinction advocacy. Read the original article →

Zero Things More Important

Yudkowsky’s political call to make AI extinction a bipartisan priority. Read the original article →

The People Who Rehearsed the Apocalypse Are Writing the Rules

AI-risk advocacy and the movement from private expertise into public power. Read the original article →

Safety: How Independent Are the Networks Behind OpenAI’s Fired Researchers?

AI-safety research, media coverage, and organizational relationships. Read the original article →

The Network Behind the Moderate

MIRI, donors, ideological networks, media amplification, and political influence. Read the original article →

The Explainer: Hank Green and the Uses of Careful Men

Science communication, institutional funding, and the authority of trusted public voices. Read the original article →

The Paperclip Problem

The funding of speculative AI risk alongside present-day surveillance and state control. Read the original article →

The Uses and Limits of Universalism

Bernie Sanders, AI extinction rhetoric, and the politics of universal versus specific harms. Read the original article →

The Safety Narrative

How AI-safety narratives acquire credibility through framing, research citations, and undisclosed affiliations. Read the original article →

What the reporting examines

The flow of funding and influence through research institutes, donors, science communicators, media outlets, and political institutions; the handling of misconduct allegations within influential communities; and the gap between speculative AI catastrophe narratives and harms already occurring.

For journalists and researchers

Each original article contains its own evidence and analysis. The existence of a relationship is not itself proof of wrongdoing or coordination. Independent scrutiny, corrections, and documentary leads are welcome.

Contact the editor →