Methodology

What we track

AI Governance Stack currently tracks 84 tools across 5 categories. A tool qualifies if it materially helps an organisation govern, monitor, constrain or test AI systems. That deliberately spans four quite different products — governance documentation platforms, observability tools, runtime guardrails and AI security testing — because buyers routinely confuse them, and the confusion is expensive.

How entries are compiled

Every entry is compiled from public sources: vendor documentation and websites, funding and acquisition records, analyst coverage, and public customer references. Each profile records what the vendor does, which frameworks it publishes explicit support for, how it deploys, which team it is built for, and its ownership status. Sources are listed at the bottom of every profile so you can check our work.

Where a fact could not be verified, we leave it blank rather than guess. A blank funding or pricing field means the information is not public — not that it does not exist. Framework support is recorded only where a vendor documents it; absence of a mapping is not evidence a tool cannot help you meet an obligation, only that you will have to evidence it yourself.

What the ordering means

Listings are ordered by how thoroughly a tool is publicly documented — a transparency proxy computed from framework mappings, published capabilities, integrations, customer references and corroborating sources. It is explicitly not a quality ranking, and it is not a recommendation. A well-documented tool can be wrong for you and a thinly-documented one can be exactly right. Use the ordering to decide what to read first, not what to buy.

How comparisons are chosen

We only compare tools within the same category, because comparing a governance documentation platform against a runtime guardrail tells you nothing useful. The vendors that get comparison pages are chosen editorially — the tools buyers actually shortlist against each other — rather than by the documentation score above. That score rewards vendors who publish long feature pages, which would otherwise crowd out the AI-governance specialists people most often compare.

Independence

No vendor pays to be listed, to be ranked higher, or to be described more favourably. There are no sponsored listings and no affiliate links to vendors. The site is published by Proceptual, which sells AI governance training — the advertising you see here is for that training, and it is labelled. We have no commercial relationship with the tools reviewed.

That does create one bias worth naming: we believe tooling alone does not produce governance, and we sell training premised on that view. Read the listings with that in mind.

Freshness

This market consolidates fast — a large share of the tools listed here have been acquired in the last two years. Every profile carries a “last reviewed” stamp, and the directory is re-reviewed on a monthly cycle that checks for ownership changes, new framework support and vendors that have shut down or been absorbed.

Corrections

If you work at one of these vendors and something here is wrong or out of date, we want to fix it. Corrections are free and do not require a commercial relationship. Send the correction with a public source and we will update the entry and its review date.