Box
Public documentation for Box AI and agent governance. Vendor statements, reviewed 4 October 2026.
01Evidence checks
The method01Customer data training statementFound
Box's AI trust page says enterprise data used by AI model providers is not used for training without prior approval.
02Model/provider disclosureUnclear
The reviewed public sources do not establish this signal for Box AI and agent governance. This is a limit of the reviewed material, not a finding about private practices.
03AI-specific privacy/data handling statementFound
Box describes AI privacy requirements, encryption and access controls.
04Subprocessor or AI vendor disclosureUnclear
The reviewed public sources do not establish this signal for Box AI and agent governance. This is a limit of the reviewed material, not a finding about private practices.
05Human oversight statementFound
Box documents content-layer controls that apply to built-in and third-party agents.
06AI usage disclosureFound
The linked documentation describes AI use in Box AI and agent governance.
07Framework references or governance alignmentUnclear
The reviewed public sources do not establish this signal for Box AI and agent governance. This is a limit of the reviewed material, not a finding about private practices.
08AI incident response statementUnclear
The reviewed public sources do not establish this signal for Box AI and agent governance. This is a limit of the reviewed material, not a finding about private practices.
09Model card/system card or equivalentUnclear
The reviewed public sources do not establish this signal for Box AI and agent governance. This is a limit of the reviewed material, not a finding about private practices.
10Responsible AI/governance policyUnclear
The reviewed public sources do not establish this signal for Box AI and agent governance. This is a limit of the reviewed material, not a finding about private practices.
11AI safety/testing/evals/red-teaming statementUnclear
The reviewed public sources do not establish this signal for Box AI and agent governance. This is a limit of the reviewed material, not a finding about private practices.
12Security/trust centre availabilityUnclear
The reviewed public sources do not establish this signal for Box AI and agent governance. This is a limit of the reviewed material, not a finding about private practices.
02 / The reading list
Original sources
03 / Your next conversation
Ask about your exact use case.
Which product and plan do these policies cover? What changes when AI features are enabled? Ask for the written answer.
Take the buyer's checklist →