Cohere Responsible Use Policy

This document is Cohere's Acceptable Use Policy (AUP) governing permissible and prohibited uses of its AI models, APIs, and related services, including the Command R and Command R+ model family. The terms establish categories of prohibited conduct, including using the services to generate content that facilitates weapons of mass destruction, CSAM, cyberattacks, disinformation, and non-consensual intimate imagery, and the agreement states that operators and users bear responsibility for ensuring downstream applications comply with these restrictions. The policy distinguishes between operator-level permissions (businesses building on the API) and end-user-level permissions, establishing a layered responsibility framework in which operators may customize permitted behaviors within bounds Cohere sets, which is operationally notable for B2B compliance teams assessing liability allocation in API deployment contexts. The document engages regulatory frameworks including the EU AI Act, which classifies certain AI applications as prohibited or high-risk, and relevant national laws governing CSAM, cyberweapons, and election interference; compliance obligations under these frameworks may constrain or supplement how the AUP terms apply in practice, depending on jurisdiction. Material compliance considerations include the policy's prohibition on use in certain high-stakes autonomous decision-making contexts without human oversight, and its explicit restriction on generating synthetic media of real persons without consent, both of which interact with emerging AI governance obligations in the EU and several US states.