AI claims should be verifiable. AI artifacts should be portable.

Prove your AI once. Use it everywhere.

Your buyers send a 300-question AI questionnaire. You answer it slowly, from memory, and differently every time — and the deal waits. We turn that into one continuously verified trust profile you can hand to every buyer, kept true as the rules change. The same discipline can make open AI easier to inspect, mirror, and choose.

One trust infrastructure. Two uses.

For organizations

Make your AI claims reusable.

Temper Services turns scattered governance evidence into a live trust profile: buyer-ready, source-linked, and kept current as the rules change.

Explore services
Working group

Make open AI independently verifiable.

Temper Commons is convening the people who need an open-AI trust layer with public rules, portable artifacts, and a credible exit path. It is an invitation to shape the work—not a claim that the work is already finished.

Explore Temper Commons

The Trust Layer

See ours →
Profile

A continuously verified trust profile

One live page per company: which frameworks you align to, which laws apply and your standing under each, with evidence dates. A status page for AI governance. Ours is public — including the frameworks we haven't earned.

Standard

55 questions, openly licensed

The canonical question set that replaces the bespoke questionnaire. Every question maps to NIST AI RMF, ISO 42001, the EU AI Act or M-25-22. Free under CC BY 4.0 — adoption is the point.

Freshness

Kept true as the rules change

SOC 2 is static; AI regulation moves weekly. When a law changes, the profiles citing it flag for re-verification. That visible movement is the product — and why a profile beats a PDF.

Regulatory Tracker

View full tracker →
US Federal

NIST AI RMF, OMB M-25-22, EO 14179

16 federal instruments — the safe-harbor framework and the procurement rules reshaping government AI contracts.

US States

Texas TRAIGA, California SB 942, NY LL 144

16 laws across the patchwork — the compliance surface most organizations feel first.

European Union

EU AI Act Article 50 — now in effect

Live since Aug 2, 2026: chatbot disclosure, deepfake labeling and emotion-recognition disclosure. Not delayed by the Digital Omnibus — if you have EU users, these obligations apply today.

Turing

A small, self-contained spiking neural network you can run and inspect in your browser — the substrate underneath everything else on this site.

See the model

Cleared & Credentialed

SAM.gov registered, TS/SCI, and NAICS codes for government AI governance work — plus downloadable compliance documents.

View credentials