Hogan Assessments

Company overview

Hogan Assessment Systems (dba Hogan Assessments) is a privately held, independent talent-assessment company founded in 1987 by Drs. Robert and Joyce Hogan and headquartered in Tulsa, Oklahoma. Its core products are the HPI (normal, 'bright-side' personality), HDS (derailing, 'dark-side' tendencies), and MVPI (motives, values, preferences), alongside the HBRI cognitive-reasoning inventory, delivered through a network of authorized distributors and partners. The company reports having assessed more than 14 million people, offers assessments in 51 languages across 66 markets, and is used by roughly 75% of Fortune 500 firms. It is grounded in decades of industrial-organizational psychology research, with a large validity archive and staff who publish in peer-reviewed journals.

Regulatory exposure

As a personality assessment used in selection and development, Hogan sits at the heart of employment-assessment regulation, yet it takes the unusual public position that its core tools fall outside the modern automated-employment-decision-tool frame. It published a statement arguing NYC Local Law 144 does not apply because humans (not computers) identify inputs and weights and because the tools are meant to be 'one consideration among many' — so no LL 144 bias audit exists. Under the EU AI Act, assessments used for selection are Annex III high-risk when they qualify as AI systems; Hogan's fixed-key psychometric scoring arguably sits on the boundary, but a separate 'AI at Hogan' page discloses it is exploring machine-learning scoring and uses NLP in backend processes, a tension with its 'we don't use AI' framing. It carries Illinois/Colorado exposure through US enterprise clients, and rests on strong traditional footings (UGESP, SIOP, APA standards, SOC 2, EU-US Data Privacy Framework) rather than AI-Act-shaped documentation.

Path to a higher score

The highest-leverage move is to reconcile the AI-scope question openly: publish an AI-system determination and, wherever ML scoring ships, an Annex IV-style model card, instructions-for-use, and explainability statement rather than a single principles page. Commission and openly publish an independent, dated bias audit (or per-year adverse-impact impact-ratio tables) instead of providing adverse-impact analyses privately to clients, and refresh the 2014 self-authored material. Add explicit EU AI Act deployer guidance and an Article 27 FRIA template, offer a self-service DPA, pursue ISO 42001/27001, and publish a data-governance and data-exclusion statement. For post-market monitoring, stand up a public status page, security.txt, and model-update changelog, and document concrete in-product oversight controls (override, audit logs, per-jurisdiction settings) beyond professional-practice guidance.