Boston Global Forum and AIWS establish an independent civil-society mechanism to observe consequential AI power, warn of emerging danger, and call those responsible to act.
Boston, September 27, 2026
The Boston Global Forum and AIWS today publish The Founding Framework of the AIWS Trust Observatory, a decisive step from warning about AI risk to building an institution that can see danger in time and move the world to act.
The Observatory is neither a global AI government nor a form of self-oversight by those who build AI. It is an independent civil-society institution for independent global observation.
Do not centralize control over AI. Do not leave consequential AI power unobserved.
How the Observatory Shapes the Future
It moves decisions earlier. Four alert levels, from Watch to Critical Alert, bring evidence of danger to those who can act while action still matters, not after harm has become irreversible.
It turns promises into accountability. The Commitment Register records every public AI safety commitment of each company and government, and measures those promises against observable conduct.
It turns blind spots into law. The Blind Spot Map states what society cannot yet see, why, and what disclosure or law would make it visible, giving policymakers concrete proposals instead of general concern.
It gives society the right to act. Every warning names who must act, what must be done, and who must answer, and the Observatory follows up until the responsible parties respond.
It makes concentrated AI power answerable. Corporate and state AI power alike are observed through Five Stations — Behavior, Footprint, Consequence, Access, and Insider — so that no one exercises consequential AI power beyond public scrutiny.
Independence as the Source of Authority
The Observatory’s authority comes from evidence, not power. Every finding carries an evidence grade. No public conclusion rests solely on confidential sources. Every subject has the right of reply and review. No one under observation may fund the Observatory.
The first Quarterly State of Frontier AI will be published at the end of January 2027.
Join the AIWS Volunteer Fellows
The Observatory invites scholars, engineers, security researchers, lawyers, journalists, and policymakers worldwide to join the AIWS Volunteer Fellows and help build humanity’s independent eyes on AI power.
AI power must never become power over society without society’s knowledge, scrutiny, and right to act.
DOWNLOAD THE FULL TEXT
→ The Founding Framework of the AIWS Trust Observatory (PDF)
