Français

Organisations

AI Whistleblower Initiative (AIWI)

Type
Civil society organisation
Lieu
États-Unis — United States
Dernière vérification
2026-09-22
Prochaine vérification
2027-03-21

Pas encore traduit — affiché en anglais.

Comment les joindre

  • Contact Hub — connects to vetted specialist attorneys · Web form
    Vérifié · 2026-09-22

    Current and former employees of AI companies

  • General contact · Web form
    Vérifié · 2026-09-22

    Anyone

  • 'Third Opinion' — independent expert assessment of whether a concern is well-founded · Written submission
    Vérifié · 2026-09-22

    Insiders unsure whether their concern warrants escalation

    Unusual and valuable: a way to get a concern sanity-checked before taking legal or public risk.

  • Homepage · Homepage
    Non contrôlé

Ce qu'il fait

Research, policy and direct support for frontier-AI employees considering disclosure. Founded by Karl Koch. Offers free consultations with specialist whistleblowing attorneys (pro bono or via a defence fund), an AI Whistleblower Defense Fund, a 'Third Opinion' service for independent expert assessment of a concern, privacy and security guidance, and a published AI Whistleblower Guide. Also does policy work on whistleblower protections for AI.

Évaluation franche

Not on the original list but it is the single most important routing destination for anyone inside an AI company with a safety concern — every other watchdog here defers to it. Small, purpose-built, responsive. UNVERIFIED: no dedicated secure email or Signal channel was published on the homepage; use the Contact Hub.

Préoccupations couvertes

Meta-level: the information asymmetry between frontier labs and the public. Agnostic as to which risk — it exists so that insiders with any credible safety concern can surface it safely.

Canal gouvernemental

Policy work on whistleblower protection statutes (e.g. proposed AI whistleblower protection legislation); no formal advisory seat.

Sources

Une erreur ?