Acceptable Use Policy

This is xAI's Acceptable Use Policy (AUP), effective January 2, 2025, with a version date of April 8, 2026. It governs all users of xAI's Service, including consumers, developers, and businesses. Its stated purpose is to "maximize your control over how you use our Service" while ensuring lawful, responsible, and safe use.

The card does not discuss model capabilities, benchmark scores, modalities, or context window.

The card does not discuss evaluation or testing methodology.

The card does not discuss red-team findings, CBRN evaluations, or catastrophic-risk assessments.

The AUP prohibits use of the Service to develop bioweapons, chemical weapons, or weapons of mass destruction, and forbids circumventing safeguards except by members of an official Red Team or those with xAI's explicit authorization. xAI reports suspected child sexual abuse material to the National Center for Missing and Exploited Children. Users can report violative content via an in-product reporting flow or by emailing support@x.ai.

The Service is available via web at grok.com, iOS, Android, and integrated into X (formerly Twitter). An enterprise tier (Grok Enterprise) and an API with voice and image generation endpoints are offered. Capitalized terms are defined in separate Consumer and Enterprise Terms of Service documents.

The card does not discuss technical limitations of the underlying model. The policy itself acknowledges that "our policies will evolve over time as our Service and user base change, as well as based on what we learn over time," implying current rules are not considered final.

The document carries a version date of April 8, 2026, against an original effective date of January 2, 2025. No specific changelog entries or version deltas are disclosed.