How We Test & Score
Every tool on this site goes through the same process before it's published, and every review is periodically re-checked against that same process to stay current. This page covers how tools are scored; see our Editorial Guidelines for our broader sourcing and independence standards. For the broader context behind why AI adoption in HR matters, see our HR AI adoption statistics.
How tools are sourced
We track new AI HR tools through market scans, reader submissions, and vendor submissions via our Contact page. Being submitted doesn't guarantee inclusion or a positive verdict.
What “Recommended,” “Worth a Look,” and “Not Yet Ready” mean
Recommended means we'd suggest this tool to a specific team profile without reservation. Worth a Look means it's solid but has a real trade-off worth knowing before you commit. Not Yet Ready means we found a gap significant enough that we can't recommend it yet — usually pricing opacity, missing core functionality, or unresolved compliance risk.
How pricing is verified
Pricing is checked directly against the vendor's current published pricing page at time of writing. Every tool profile shows a “last verified” date — pricing changes frequently, so always confirm current numbers directly with the vendor before purchasing. When a vendor doesn't publish a price at all (“Custom” or “Contact sales”), we say so rather than guessing — we never invent a number or display “Free” for a tool that's actually enterprise-quoted.
How ratings are sourced
A rating shown on this site is one of two things, and every tool page makes clear which: either AI HR Atlas's own editorial assessment, based on the same criteria described above, or a rating pulled from a third-party platform like G2 — in which case it's explicitly labeled with the source and never presented as our own judgment. We don't blend the two into a single unlabeled number.
How deep our research goes
Today, every tool on this site is evaluated through structured desk research — vendor documentation, published pricing pages, product screenshots, and public review data — rather than a hands-on trial account. This is disclosed per tool rather than implied away, and it's a deliberate honesty choice: we'd rather say “desk research” than imply first-hand testing that didn't happen. If that changes for a specific tool, the page will say so. See our AI Transparency Policy for how AI is used in producing this content in the first place.
How we treat vendor claims
Product facts (pricing, features, integrations) are sourced from the vendor's own documentation and pricing pages, not marketing copy or press releases, and dated so you can judge how current they are. A vendor publishing a claim doesn't make it part of our verdict — the Recommended / Worth a Look / Not Yet Ready tiers above reflect our own evaluation criteria, independent of how a vendor markets itself.
Corrections
If you spot outdated pricing, a factual error, or something that's changed since we last checked, let us know— we review every correction request and update the page when it's warranted.
How we make money — and what that doesn't affect
This site earns revenue through affiliate partnerships, sponsorships, and (eventually) vendor advertising. None of that determines a verdict. A tool with no affiliate relationship can be marked Recommended, and a sponsor's tool can be marked Not Yet Ready — see our Affiliate Disclosure for full detail.