We count visits with Google Analytics cookies and Vercel Web Analytics to see which pages help readers most.

Cookie settings
The Assurance Desk

BABL AI Inc. review

Advisory firm

A US AI-audit firm running New York City Local Law 144 bias audits and ISO 42001 and EU AI Act assessments. Its named clients are hiring-technology companies.

Relevant if you sell hiring software and need a New York City Local Law 144 bias audit

Price Published
Indemnity cover Not published
How it uses AI Published
Facts last checked

Not published means we did not find it on the firm's own pages.

BABL AI scores

Six scores, each 1 to 5. The evidence for each is further down.

Help when things go wrong 3/5
Qualifications of the people who do the work 3/5
What it publishes about fees 3/5
Knowledge of your sector 4/5
Independence from what it sells 5/5
Who carries the risk if the advice is wrong 3/5

What BABL AI offers, and what it publishes

BABL AI runs New York City Local Law 144 bias audits, or checks a client's own testing, and does ISO 42001 gap analyses and US NIST AI RMF and EU AI Act assessments. It says it works to the ISAE 3000 assurance standard and that a single-system audit typically takes five to eight weeks.

BABL AI does not publish a price for its audits; its AI audit training programme costs $2,500. It says the audit fee is fixed in the engagement letter and never depends on the findings. It does not publish indemnity cover, and its public list of vendors handling client data is empty.

Four named clients, iCIMS, BrightHire, HackerRank and RippleMatch, give on-the-record testimonials about completed audits. It says it does not design, build or fix the systems it audits. Its privacy policy says its work is always initiated, directed and reviewed by people. Its team page names 14 people but gives no qualifications. We could not confirm its registration in Iowa.

Strengths

  • Runs NYC Local Law 144 bias audits directly (attestation or independent testing), plus ISO 42001 gap analysis and NIST AI RMF/EU AI Act-aligned assessment work, as named service lines.
  • Four named, on-the-record clients (iCIMS, BrightHire, HackerRank, RippleMatch) give attributed testimonials naming specific completed algorithmic-bias audits, including a claimed 'first-ever bias audit report in interview intelligence' for BrightHire.
  • States a fixed, non-contingent fee model and says it does not design, build or remediate the systems it audits; advisory and assurance work are staffed and disclosed separately, mirroring ISAE 3000 independence rules used in financial audit.
  • Has operated as a dedicated AI-audit specialist since 2018, longer than NYC LL144, the EU AI Act or the Colorado AI Act have existed.
  • Embedded in the AI-audit standards world: founder Shea Brown holds a ForHumanity board seat, the firm is a founding member of the International Association of Algorithmic Auditors, and it sits on the PACT AI Leadership Council.

Things to check

  • The team page names 14 people and titles but publishes no individual degrees, professional certifications or prior employers for any of them.
  • The public subprocessors page promises disclosure but its 'current providers' table is empty, so the actual vendors handling client data are not named there.
  • No fee figures or rate card are published anywhere on the site, only the fee structure (fixed, non-contingent) is disclosed.
  • We could not read the state register. Ask for its company details and current status.

Common questions

Does BABL AI run New York City Local Law 144 bias audits?

Yes. It runs them itself or verifies a client's own testing.

How does BABL AI charge for an audit?

It says it charges a fixed fee, set in the engagement letter, that never depends on the findings. It publishes no audit price; its training programme costs $2,500.

Is BABL AI's US registration confirmed?

No. We could not confirm its status on the Iowa Secretary of State's register or through OpenCorporates when we checked on 20 September 2026. This is a gap, not a finding either way.

Does BABL AI say how it uses AI?

Yes. Its privacy policy says its work is always human initiated, directed and reviewed, and that it makes no solely automated decisions with legal effect.

Evidence for each BABL AI score

Show our working

Notes behind each thing to check

  • The team page names 14 people and titles but publishes no individual degrees, professional certifications or prior employers for any of them. (babl.ai/who-we-are/ fetched 2026-09-20 lists names and titles only, no bios or credentials.)
  • The public subprocessors page promises disclosure but its 'current providers' table is empty, so the actual vendors handling client data are not named there. (babl.ai/subprocessors/ fetched 2026-09-20 shows an empty current-providers table.)
  • No fee figures or rate card are published anywhere on the site, only the fee structure (fixed, non-contingent) is disclosed. (babl.ai/faq/ fetched 2026-09-20 states the structure only, no dollar figures.)
  • We could not read the state register. Ask for its company details and current status. (Iowa SOS and OpenCorporates both failed via direct fetch and Bright Data fallback on 2026-09-20.)
Six criteria, each scored 1 to 5, with the evidence for each.
CriterionScoreEvidence
Help when things go wrong 3/5 Runs LL144 audits and covers NIST AI RMF/EU AI Act per its service page, but no named regulatory-enforcement defence or litigated-incident work was found in this pass.
Qualifications of the people who do the work 3/5 No individual staff degrees or certifications are published, but the firm holds a ForHumanity board seat and a founding IAAA membership at the org level.
What it publishes about fees 3/5 FAQ states fees are fixed in the engagement letter and never contingent on findings, but no rate card or fee range is published anywhere.
Knowledge of your sector 4/5 Four named, quoted clients (iCIMS, BrightHire, HackerRank, RippleMatch) are all HR-tech hiring-algorithm bias audits, a specific and well-evidenced specialism.
Independence from what it sells 5/5 FAQ states BABL AI does not design, build or remediate the systems it audits, and separates advisory staffing from assurance staffing per ISAE 3000 rules.
Who carries the risk if the advice is wrong 3/5 Uses ISAE 3000-based reporting with senior-advisor quality review before issuance, but no PI-insurance or liability-cap disclosure was found on the site.

Sources for BABL AI

  • BABL AI was founded in 2018 as an independent AI-audit firm, before most AI-governance law existed. (Founded 2018) (babl.ai, checked 20 September 2026)
  • BABL AI runs NYC Local Law 144 bias audits directly, or verifies a client's own testing, as a named service line. (NYC LL144 bias audits offered) (babl.ai, checked 20 September 2026)
  • BABL AI states its fee is fixed in the engagement letter and is never contingent on audit findings. (Fixed, non-contingent fee model) (babl.ai, checked 20 September 2026)
  • BABL AI runs audits to the ISAE 3000 assurance standard and typically completes a single-system audit in five to eight weeks. (ISAE 3000 methodology, 5-8 week timeline) (babl.ai, checked 20 September 2026)
  • BABL AI's privacy policy states its work is always human initiated, directed and reviewed, and it does not use solely automated decisions with legal effect. (Discloses own AI-use boundary) (babl.ai, checked 20 September 2026)
  • Four named clients, iCIMS, BrightHire, HackerRank and RippleMatch, give attributed, on-the-record testimonials about completed bias/risk audits. (4 named, quoted HR-tech audit clients) (babl.ai, checked 20 September 2026)
  • BABL AI publishes a price for its AI Audit & Assurance Training Program: $2,500. It publishes no figure for its audits. ($2,500 (training programme)) (babl.ai, checked 29 September 2026)

Seeking ISO 42001 certificationExposed to the EU AI ActRegulated sectors

Where this firm appears

If you represent this firm and something here is wrong or out of date, email hello@theassurancedesk.com with the evidence.

This is an independent editorial guide, not legal advice.