No model training by defaultOn Team and Enterprise, Anthropic does not train models on your content by default; on individual plans it is a setting you can opt out of.
Source: claude.com/pricingAI policy and governanceAI policy, security and governance for Claude
Rules people follow because they are short and match how the tools work. For a small business that is a one-page policy and the right settings; for a large organization it is a governance model tied to identity, logging and retention.
Who it fits
Works on: Any plan; admin and compliance controls on Team and Enterprise.
Solo owners and individuals
Get a working setup in one sitting: the right individual plan, your apps connected, a few Projects and Skills for the jobs you repeat every week.
Usual plan: Pro, Max or UltraSmall teams
One shared workspace with central billing and single sign-on, shared Projects, a short staff workshop and a one-page AI use policy people actually read.
Usual plan: TeamMid-size companies
Role-based training by department, connectors to your CRM, files and field systems, organization-wide Skills, and usage analytics to see what is being adopted.
Usual plan: Team, or Enterprise when you need SCIM, audit logs or custom retentionLarge organizations
A governed rollout: SSO and SCIM, role-based access, audit logs and the Compliance API wired into your controls, a champions network, and Claude Code for engineering.
Usual plan: EnterpriseWhat we do
- A plain-language AI use policy: approved tools, what data may be shared, when a person must review output.
- Settings checked against the policy, including whether your content may be used for model training.
- Connector rules: which apps are allowed and what each role can reach.
- On Team and Enterprise, admin controls and organization instructions configured to match.
- On Enterprise, audit logs, the Compliance API and retention settings mapped to your security and records requirements.
- A short training module so everyone knows the rules and why they exist.
What you have at the end
- 1A policy people read
- 2Claude settings that match it
- 3Approved connectors by role
- 4Logging and retention where you need them
The Claude features this uses
As Anthropic describes them; each card links to the source.
No model training by defaultSingle sign-on and central adminSSO, domain verification, central billing and admin controls for connectors on Team and Enterprise.
Source: claude.com/pricingEnterprise controlsRole-based access, SCIM, audit logs, the Compliance API, custom data retention and domain capture on Enterprise.
Source: claude.com/pricingHow it runs
The same stages, in this order. For a small business several happen in one session.
- 1Demo and discovery
A 15-minute demo on work like yours, then a short conversation about who would use Claude, for what, and with which systems.
- 2Pick the plan and the first uses
The plan that fits your size and the two or three jobs that pay back first. Small businesses usually stop here and go straight to setup.
- 3Set up and connect
Accounts, and for teams single sign-on and admin settings; connectors and the Business Suite MCP connection; a one-page AI use policy.
- 4Train on your own work
Hands-on sessions built from your tasks, with role-based tracks for larger teams and Claude Code sessions for engineers.
- 5Build Projects, Skills and workflows
The repeatable work becomes Projects and Skills, connected to your systems and automations where it saves a step.
- 6Measure and hand over
What people use and where they get stuck, a library your team owns, and office hours for questions afterwards.
Frequently asked questions
Does a small business really need an AI policy?
A short one, yes. One page that says which tools are approved and what customer or financial data never goes in prevents most problems, and it costs little to write.
Does Anthropic train on our data?
On Team and Enterprise, Anthropic's pricing page says your content is not used for model training by default. On individual plans it is a setting; we check it with you during setup.
Can you help with a regulated business?
We map Claude's controls to your requirements and work with your compliance lead. Anthropic lists a HIPAA-ready offering on Enterprise; whether it fits your obligations is a decision for you and Anthropic.
Consultants in a Box is an independent consultancy; this page is ours, not published or reviewed by Anthropic. Claude and Anthropic are trademarks of Anthropic PBC; other names and logos belong to their owners. Plan and feature facts are from Anthropic's pages linked above, reviewed November 6, 2026.
Talk about ai policy and governance
Book a 15-minute demo. We will show you Claude on work like yours and the first step that fits your size.