Software vendor

Anthropic

Anthropic contracts as Anthropic, PBC outside the EEA, Switzerland and the UK, and sells Claude Team and Enterprise, with SCIM and audit logs on Enterprise.

1 offering in the directory Last verified October 21, 2026

Function

Across functions

Ownership: Not stated
Not stated in the sources read. The Commercial Terms name Anthropic, PBC (and Anthropic Ireland, Limited for EEA, Swiss and UK customers) as the contracting entity.
Headquarters: Not stated
SEC reporting: No
Audit firm: No
Website: anthropic.com
Subprocessor list: Not stated
Contract terms: anthropic.com/legal/commercial-terms

Offering · Software

Claude Team and Enterprise plans

Directory recordLast verified October 21, 2026Compare with othersBuild questions for this vendor

Anthropic sells Claude to organizations on a Team plan for 2 to 150 people and an Enterprise plan, both including Claude Code, Claude Cowork, enterprise search and single sign-on.

Anthropic's published pricing, October 2026, lists Team at $20 per standard seat per month billed annually and Enterprise at US$20 per seat per month with a 20-seat minimum plus usage at API rates.

Role-based access, SCIM, audit logs, a Compliance API and custom retention controls are Enterprise-only.

Anthropic's Commercial Terms say it may not train models on customer content, while its Privacy Center says feedback a user submits may be used for training.

Where it fits

CoE analysis Fits organizations that want a general assistant together with Claude Code and Claude Cowork in one subscription, with the Enterprise plan for identity and audit controls.

CoE analysis Enterprise cost depends on usage at API rates as well as seats, so a usage forecast and spend limits belong in the business case.

What it does

Offering type: Software
Workflow areas: Enterprise assistant
Enterprise search
Customer size: Not stated

AI

Kind of AI: General-purpose assistant
Suggests or acts: Not stated
Availability: Generally available
Team and Enterprise are offered on the published pricing page; Claude Motion and Dashboards are labeled beta, and Claude Security (beta) is listed for Enterprise.
Models named: Anthropic

Control

Who controls it: Customer IT

Commercial

Pricing model: Published list price
Public price: USD 20 seat/month, billed annually
Term: annual

Evidence

Independent outcomes: None found in the sources read.
Sources read: Vendor-stated 8
Independent market evidence None read
Independently verified outcomes None read
Pricing 2
Known limitations None read
Coverage: Directory record
Last verified: October 21, 2026

Security and control

Single sign-on: Supported
Single sign-on is included in Team, Enterprise (self-serve) and Enterprise (sales-assisted). Protocols are not named on the pricing page.
User provisioning: Some plans
Protocols: SCIM
SCIM is Enterprise-only; not included in Team.
Role-based access: Supported
Role-based access with fine-grained permissioning on Enterprise only.
Administrator controls for AI: Not stated
Admin controls for remote and local connectors (Team and Enterprise); user and organization spend limits (Enterprise); owners can disable feedback submission with the Rate chats setting.
Audit logs: Exportable: Yes
Enterprise only. Owners can export the past 180 days of audit logs; chat and project titles and contents are excluded (identifiers only), and chat inputs and outputs will be exportable by Primary Owners through data exports. The Export logs button cannot be used with customer-managed encryption keys; audit log events are also in the Compliance API. Whether logs record actions taken by Claude Code or Cowork is not stated.
Certifications: SOC 2 Type 2
SOC 2 Type I and Type II listed for commercial products (Claude for Work and the API); product scope and period not stated.
ISO 27001
ISO 27001:2022 listed for commercial products.
ISO 42001
ISO/IEC 42001:2023 (AI management systems) listed for commercial products.
Other
HIPAA-ready configuration with a BAA available, listed for commercial products.
Data retention: Prompts and outputs: Chats are retained in the product; a deleted conversation leaves chat history immediately and back-end storage within 30 days. Chats flagged for Usage Policy violations are kept up to 2 years and classification scores up to 7 years; feedback submissions are kept up to 5 years. Anthropic may retain chats or sessions as required by law or as necessary to combat Usage Policy violations, and Covered Models carry limited retention and review requirements. Enterprise has custom data retention controls.
Logs: Audit log export covers the past 180 days.
Users delete chats themselves; back-end deletion within 30 days.
Data residency: Not stated
Not stated on the pages read.
Training on customer data: Not used
Scope: Vendor models
Where stated: Contract
Commercial Terms: Anthropic may not train models on Customer Content from Services. The Privacy Center adds that feedback a user explicitly submits may be used for training; see conflicts.

AI features

As the vendor or firm describes them. Notes in small print record what the sources leave open.

Claude chat and enterprise search
Claude with enterprise search across the organization and connectors managed by admins.
Kind
General-purpose assistant
Suggests or acts
Suggests
Who controls it
Customer IT
Availability
Generally available
Control: Admin controls for remote and local connectors on Team and Enterprise.
Claude Code and Claude Cowork
Included in the Team and Enterprise plans; the pages read name these products but do not describe what they do. A news line on the pricing page says Claude Cowork is now just Claude, rolling out to Pro and Max first.
Kind
Not stated
Suggests or acts
Not stated
Who controls it
Customer IT
Availability
Generally available
Control: Enterprise admins set user and organization spend limits. Kind of AI not stated in the sources read.

Unclear or conflicting statements

Where the sources are unclear or disagree, both statements are shown with their links.

Use of commercial customer content for model training

Statement A

Commercial Terms of Service (effective June 17, 2025): Anthropic may not train models on Customer Content from Services.

anthropic.com/legal/commercial-terms

Statement B

Privacy Center article for commercial products (August 18, 2026): by default inputs and outputs are not used for training, but if a user explicitly reports feedback or otherwise allows it, Anthropic may use those chats and coding sessions to train its models.

privacy.claude.com/en/articles/7996868-is-my-data-used-for-model-training

Recorded under D-143; not resolved. Team and Enterprise owners can disable feedback with the Rate chats setting [C-30656]. The Terms also say, in section A.3, that Anthropic may use feedback the customer provides [C-31105], which may bridge the two statements.

Both statements are published as found; CoE draws no inference about which governs.

Questions to ask

For this offering

Which of our users can submit feedback that Anthropic may use for training, and have we turned off the Rate chats setting where needed?
How will usage at API rates on the Enterprise plan be forecast and capped by user and by organization?
What do the audit logs and the Compliance API record for agentic work done through Claude Code and Claude Cowork?
Does Anthropic's SOC 2 Type II report and ISO/IEC 42001 certificate cover the Claude apps we would deploy, and for what period?
Where will our chats and files be stored and processed, given that the pages read do not state data residency?

From the diligence question set

Build questions for this vendor