UKG

Company overview

UKG (Ultimate Kronos Group) was formed on April 1, 2020 from the merger of Kronos Incorporated (founded 1977) and Ultimate Software (founded 1990), with dual U.S. headquarters in Lowell, Massachusetts and Weston, Florida. It is one of the largest human-capital-management and workforce-management vendors, serving thousands of enterprise customers. Within its UKG Pro suite, UKG Pro Recruiting provides applicant tracking, and the Bryte AI layer (launched 2023) adds generative AI assistance and profile-match candidate scoring. UKG positions Bryte as 'fair, inclusive, and secure by design,' trained on a large workforce dataset. As an enterprise SaaS HCM vendor sold via custom contracts, UKG sits firmly in the enterprise price tier.

Regulatory exposure

UKG's recruiting AI plausibly falls within scope of high-risk employment-AI regimes: candidate match scoring used to rank or shortlist applicants is the kind of tool covered by NYC Local Law 144 (AEDT) and is classified high-risk under Annex III of the EU AI Act, with parallel exposure under Illinois and Colorado law. UKG maintains an active responsible-AI program and an EU AI Act statement, and acknowledges the high-risk classification of hiring tools in its public HR-leader blog, but its concrete compliance artifacts (Responsible AI Statement, Bryte AI model-training documentation, EU AI Act statement) are gated behind its SafeBase Trust Center request flow rather than freely downloadable. No public NYC LL 144 third-party bias audit, no ISO 42001 certification, and no Article 27 FRIA template could be located. Because UKG is the AI provider while its employer customers are the deployers who must commission bias audits and FRIAs, the absence of freely available audit evidence shifts substantial verification burden onto deployers.

Path to a higher score

UKG could materially raise its score by ungating its responsible-AI artifacts: publish the Responsible AI Statement and Bryte AI model-training/data-governance documentation as open web pages or downloadable PDFs rather than behind the Trust Center request wall, and add an explainability/system-card document for the candidate-scoring model (Article 11-style instructions for use). The single highest-impact move is commissioning and publicly posting an independent NYC LL 144 bias audit of the recruiting match-scoring tool, ideally renewed annually, as peer ATS/matching vendors have done. Publishing an Article 27 FRIA template and explicit deployer-obligation guidance, and pursuing ISO 42001 certification, would close the remaining gaps.