BrightHire

Company overview

BrightHire is a New York-based interview intelligence company founded in 2019 by Ben Sesser and Teddy Chestnut, backed by roughly $36M from Index Ventures, 01 Advisors and Flybridge, and used by 450+ companies. Its original product records, transcribes and analyzes live interviews (originally built on Zoom), producing competency-structured AI notes that map to and auto-fill ATS scorecard fields, plus interviewer coaching and process analytics. It has since added BrightHire Screen, a voice-based AI interviewer that conducts first-round screens and scores responses against recruiter-defined rubrics, with candidate ranking and sorting in the reviewer interface. Pricing is not public: the site lists Recruiters, Teams and Enterprises tiers, all demo/quote-led, with Screen sold standalone or bundled.

Regulatory exposure

BrightHire is materially more than a notetaker. BrightHire Screen scores and ranks candidates against rubrics, and the core product generates AI competency assessments that flow into ATS scorecards, which places both inside the AEDT definition in NYC Local Law 144 and Annex III(4) of the EU AI Act, whose high-risk obligations apply from 2 August 2026. Recording and analyzing video interviews also engages the Illinois AI Video Interview Act, while employer-deployers carry Colorado SB 24-205 impact assessments and Illinois HB 3773 duties. BrightHire's declared posture is deployer-supporting rather than provider-conformant: its trust center flags EU AI Act at 'na' maturity with the single note that BrightHire 'supports its clients in their compliance with the EU AI Act.' It has commissioned annual independent bias audits from BABL AI, notably with separate audits scoped to Interview Intelligence and to Screen, and it publishes an unusually complete DPA with a 23-entry subprocessor list. The weakness is publication: every audit, AI governance policy and risk-management document sits behind a gated SafeBase portal, with no impact ratios, audit dates, LL 144 summary, or jurisdiction-specific deployer guidance anywhere public.

Path to a higher score

The largest gains are publication rather than new work. Releasing the BABL AI bias audits, or at minimum dated per-product summaries with selection rates and impact ratios plus an NYC LL 144 summary page, would move bias-audit transparency sharply. Ungating the AI Governance Policy, AI Risk Management and AI FAQ documents and adding a system/model card and instructions-for-use for Screen — covering intended purpose, rubric scoring mechanics, known limitations and accuracy metrics — would lift Article 11. An Article 27 FRIA template plus per-jurisdiction deployer guidance for NYC, Illinois, Colorado and the EU would raise both FRIA support and customer documentation. ISO 42001 certification, a published AI model-update changelog, an AI incident-reporting channel and a dedicated security@ mailbox would close the remaining post-market and technical-documentation gaps.