Claude, Anthropic’s generative AI assistant, is available for purchase through authorized university purchasing methods while UCLA works toward establishing an enterprise-grade, university-managed Claude environment. These purchasing options are only available for approved use cases not to exceed data sharing above Protection Level 2 (PDF).
Purchasing Options for Faculty and Staff
Faculty and staff may use a university PCard for individual Claude subscriptions (e.g., Pro or Max) and Anthropic API/token charges, subject to university purchasing and PCard policies. PCards should be used for recurring charges. Prior to payment, a Finance Support Portal case with an attached, approved TPRM is required to open the software Merchant Category Code (MCC) 5734. MCC requests submitted without an approved TPRM cannot be processed. If your department does not have a PCard you can request one through the Finance Support Portal. Please note that Claude Team and Enterprise subscriptions are currently excluded for purchase on PCard.
Reimbursement requests for purchases made using personal funds will be evaluated on a case-by-case basis, subject to applicable university policies, departmental approval, and funding availability. Reimbursement is not guaranteed, and faculty and staff are encouraged to use authorized university purchasing methods whenever possible.
For step-by-step instructions on submitting a reimbursement request, including documentation requirements, expense report submission, and approval procedures, please refer to UCLA Purchasing & Accounts Payable’s Lifecycle of a Non-T&E Reimbursement. Reimbursement requests must follow applicable university requirements and submission timelines.
These interim purchasing options are intended primarily for individual faculty, staff, and researcher use in support of university business, academic activities, and research. Departments and IT units considering Anthropic services for departmental applications, shared services, production systems, or integrations involving university data should consult with the UCLA Governance Risk and Compliance team (GRC) before purchasing or deploying these solutions as a robust catalog of enterprise AI offerings exist to build scalable AI solutions.
PCard authorization does not constitute approval to deploy departmental or enterprise applications. Such implementations may require additional security, privacy, architecture, and third-party risk assessments.
University-Managed Claude Environment
UCLA is working toward establishing a university-managed Claude environment in close partnership with the UC System to provide enhanced security, centralized administration, and institutional support.
As this environment becomes available, the university will provide additional guidance regarding access, purchasing, and any applicable migration requirements for existing Claude users. Faculty, staff, and researchers should consider potential changes to service availability and administration when making purchasing decisions.
Support and Service Limitations
Individually purchased Anthropic services do not include university-managed support through Digital & Technology Solutions (DTS) or centralized administration. Support for these services will be provided on a best-effort basis, subject to available university resources and Anthropic's standard support offerings.
Faculty, staff, and researchers should consider these limitations before relying on individually purchased Anthropic services for business-critical activities or operational applications.
Third-Party Risk Management (TPRM)
A Third-Party Risk Management (TPRM) assessment is currently required before purchasing Claude services. This requirement helps ensure that the use of Anthropic services aligns with applicable university security, privacy, and compliance requirements.
The assessment process may be updated as UCLA expands access to Claude and establishes additional institutional purchasing and approval pathways. Any changes to these requirements will be communicated through appropriate university channels.
Data Protection
Anthropic products are limited to P1 (Public) and P2 (Internal) data (PDF), subject to applicable university approvals and security requirements.
Do not enter P3 or P4 data, or other sensitive, confidential, or restricted university information, unless the specific Claude service has been explicitly approved for that type of data. Faculty, staff, and researchers are responsible for ensuring that information submitted to AI services complies with university data classification, privacy, and information security requirements.
AI Tools Comparison
For information about available AI services, university-managed offerings, purchasing options, and approved data protection levels, faculty, staff, and researchers may refer to UCLA's Campus AI Tools Comparison.
Commercial AI products purchased independently do not automatically receive the same security protections, administrative controls, or enterprise support as UCLA-managed services. Information submitted to these services may be retained or used by the vendor to train or improve its AI models or for marketing purposes, depending on the applicable terms of service, privacy settings, and data protection agreements.
Note: UCLA-managed flagship AI offerings (institutionally approved and supported AI platforms with established contractual protections for data privacy, security, confidentiality, and data usage) from Google, OpenAI, and Microsoft are available through Software Central. Non-flagship AI tools (non-enterprise AI solutions with limited institutional contractual protections), including individually purchased Anthropic subscriptions and API services, are restricted to Protection Levels P1 (Public) and P2 (Internal) and require an approved TPRM.