Achievements
Since December 2024: a 356-page Strategic Memo v2.6 on the paths to a proper US-China-led AI treaty; the most detailed profiles anywhere (150 pages, 900 sources) of the key humanist US AI policy powerbrokers that could jointly persuade President Trump and why. And the first engagements with their senior staff. All on about $140,000 and 2,100+ volunteer hours.
In February 2025 we saw what few did: the decisive task was to identify, profile, engage and persuade a critical mass of those power brokers, on two bets:
● First, that only an extraordinarily bold, timely and proper Trump–Xi treaty could make AI's exploding capabilities safe and fair.
● Second, that China would be willing to negotiate one.
Momentum
● Strategic Memo v3.0, about 200 pages, due November 20: a two-page brief for the President, updated AI-governance profiles of those powerbrokers, and the seven AI predictions behind every position.
● 95+ people close to those powerbrokers engaged one-to-one, including 23+ AI-lab officials and 28+ DC-based AI-governance experts, many during two US tours.
● 30 hours with key senior AI staff reporting directly to four of the twelve identified key US AI policy powerbrokers — including about a dozen face-to-face one-to-ones.
● Closed-door consultations, in person and remote, with experts, advisors and delegates of those powerbrokers. Larger ones set provisionally for December 9 and 13, January 13 and February 10.
Timeline
● 2014–2024 — Roots. Our convener, the Trustless Computing Association, fostered a new international certification body for ultra-critical IT, the Trustless Computing Certification Body, and ran nine editions of Free and Safe in Cyberspace on three continents. In 2023 it launched the Harnessing AI Risk Initiative, the Coalition's precursor.
● Dec 18, 2024 — Launch. Open Call to "President-Elect Trump and President Xi"; a 90-page Case for a Coalition with 22 contributors; six founding NGOs, later ten.
● Feb 2025 — The pivot. From a multilateral coalition of the willing to Trump and his AI powerbrokers.
● Sep 2025 — Memo v2.0. 133 pages on the feasibility of a US-China-led treaty.
● Oct 2025 — Washington and the Bay Area. Three weeks of meetings; three private dinners.
● Dec 2025 — Memo v2.6. 356 pages, 667+ sources, 150 pages of powerbroker profiles, with 24 contributors.
● Q1 2026 — Vatican track. Extensive engagement with Vatican AI leaders; the planned Rome convening merged into Washington.
● Q2 2026 — Foundations. US tax-deductible giving via a 501(c)(3) fiscal sponsor; Memo v2.8 Executive Summary; Open Letters to President Trump (1, 2, 3); our Open Case for Anthropic, which argued that verification is infrastructure to build now, not a precondition.
● Q3 2026 — Toward the table. Memo v2.9 (Aug 6); an updated Open Case for Anthropic and adjacent funders; a September week of meetings in Washington.
● Q4 2026 — After the Trump–Xi summit. Our post Cornered maps the scenario Washington must avoid. A pre-publication draft of Memo 3.0, for publication on November 20 ahead of the G20 at Doral, with Consultations set provisionally for December 9 (Washington), December 13 (Miami), January 13 and February 10 (Washington).
Resources
Mostly funded by leading AI experts and scientists, including Steve Omohundro, David Krueger, Tsvi Benson-Tilsen, Ryan Kidd and others. We have run so far on $7,500 a month. Our existing access creates opportunities; funding would let us test whether sustained follow-through produces policy use. Donate to us.
(Last updated October 29, 2026.)