Get the weekly research letter
Companies change their terms quietly. We read every version and catch what actually changed. One email a week on the changes that matter and what they mean. No account.
The document discloses that Command R models capture historical social biases and explicitly states they should not be used to make decisions about individuals or groups, citing CV ranking as a specific dangerous application.
This analysis describes what Cohere's agreement states, permits, or reserves. It does not constitute a legal determination about enforceability. Regulatory applicability and practical outcomes may vary by jurisdiction, enforcement context, and individual circumstances. Read our methodology
This provision identifies a concrete class of prohibited deployment patterns, specifically automated or AI-assisted individual assessment in hiring and similar contexts, grounded in documented bias risks, which intersects with anti-discrimination law and emerging AI governance requirements in multiple jurisdictions.
The document states that Command R models should not be used to make decisions about individuals or groups due to embedded historical social biases. The document specifically identifies CV ranking systems as a dangerous application of model outputs.
Cross-platform context
See how other platforms handle Historical Social Bias Disclosure and Use Restriction and similar clauses.
Compare across platforms →Monitoring
Cohere has changed this document before.
Receive same-day alerts, structured change summaries, and monitoring for up to 25 platforms.
"Language models capture problematic associations and stereotypes that are prominent on the internet and society at large. They should not be used to make decisions about individuals or the groups they belong to. For example, it can be dangerous to use Generation model outputs in CV ranking systems due to known biases (Nadeem et al., 2020).Excerpt from Cohere's Responsible Use Policy
1. REGULATORY LANDSCAPE: The use of AI systems with documented bias in employment screening decisions engages Title VII of the Civil Rights Act in the US, the EU AI Act's high-risk AI system classification for employment-related AI, and the EEOC's guidance on AI and employment discrimination. New York City Local Law 144 requires bias audits for automated employment decision tools used in hiring within NYC, directly relevant to the CV ranking example cited in the document. 2. GOVERNANCE EXPOSURE: High for customers deploying Command R in any employment screening, CV ranking, or candidate assessment workflow. The document's explicit citation of CV ranking as a dangerous application, combined with the known bias research it references, creates significant due diligence obligations for HR technology vendors and enterprise HR teams using the models. 3. JURISDICTION FLAGS: New York City's automated employment decision tool bias audit requirement applies to tools used in hiring within NYC. The EU AI Act classifies AI systems used in recruitment and employee selection as high-risk. Illinois AI Video Interview Act requires disclosure and consent for AI analysis of video interviews. These jurisdictional obligations are independent of Cohere's model card recommendations. 4. CONTRACT AND VENDOR IMPLICATIONS: HR technology vendors and enterprise customers using Command R in employment workflows should conduct independent bias audits and document compliance with applicable anti-discrimination and AI governance requirements. Procurement contracts should specify responsibility allocation for bias monitoring and incident response. The model card's explicit warning about CV ranking creates constructive notice of the risk. 5. COMPLIANCE CONSIDERATIONS: Legal and HR teams should audit any existing or planned workflows that use Command R outputs to inform individual employment, housing, or credit decisions. Bias testing documentation should be created and maintained for any such deployment. Where NYC Local Law 144, the EU AI Act, or similar frameworks apply, formal compliance procedures including bias audits, disclosures, and candidate notifications may be required.
Full institutional analysis
Regulatory citations, enforcement risk, and due diligence action items.
Monitor: same-day alerts on the platforms you choose. Analyst: full institutional analysis.
Compliance Governance Intelligence
Need to monitor specific governance provisions?
Compliance includes provision-level monitoring, governance timelines, regulatory mapping, and audit-ready analysis.
Built from archived source documents, structured governance mappings, and historical version tracking.
This provision identifies a concrete class of prohibited deployment patterns, specifically automated or AI-assisted individual assessment in hiring and similar contexts, grounded in documented bias risks, which intersects with anti-discrimination law and emerging AI governance requirements in multiple jurisdictions.
The document states that Command R models should not be used to make decisions about individuals or groups due to embedded historical social biases. The document specifically identifies CV ranking systems as a dangerous application of model outputs.
No. ConductAtlas is an independent monitoring service. We are not affiliated with, endorsed by, or sponsored by Cohere.