Does Claude train on your data? Free vs Pro vs Enterprise
Claude has different data terms on personal and business plans. Personal: Free, Pro, Max. Business: Claude for Work (Team / Enterprise). Every value links to Anthropic's own page. Last verified .
Claude (Free / Pro / Max)
You choose at signup
You choose at signup whether your chats are used for training.
Full personal-plan page →Claude for Work (Team / Enterprise)
Doesn't train
Doesn't train on business data by default.
Full business-plan page →Side by side
| What we checked | Personal Claude (Free / Pro / Max) | Business Claude for Work (Team / Enterprise) |
|---|---|---|
| Trains on your data by default | You choose at signup Since the Aug 28 2025 consumer-terms update, Free/Pro/Max users must choose whether chats are used for training, at signup or in a required pop-up. Anthropic's pages checked do not say which position the toggle starts in. Chats flagged for safety review can be used for safety-classifier training regardless. Source: anthropic.com/news/updates-to-our-consumer-terms (checked Oct 17, 2026), privacy.claude.com/en/articles/10023580-is-my-data-used-for-model-training (checked Oct 17, 2026) | No Anthropic does not use inputs/outputs from commercial products (Claude for Work, API) to train models, unless you submit feedback or opt in. Source: privacy.anthropic.com/en/articles/7996868-is-my-data-used-for-model-training (checked Oct 17, 2026) |
| Can you turn training off? | Yes Settings > Privacy > "Help improve our AI models" toggle. Turning it off stops future training on new and stored chats, but data already in training runs or trained models stays. Source: privacy.claude.com/en/articles/12109829-how-do-i-change-my-model-improvement-privacy-settings (checked Oct 17, 2026) | Not tracked for business plans |
| Data retention | 30 days after deletion; up to 5 years if training is on Training off: deleted chats leave back-end storage within 30 days. Training on: de-identified data may stay in training pipelines up to 5 years. Feedback (thumbs) kept 5 years. Source: privacy.claude.com/en/articles/10023548-how-long-do-you-store-my-data (checked Oct 17, 2026) | Unknown Retention page checked describes API defaults; Claude for Work retention not stated there. Pages checked: privacy.anthropic.com/en/articles/7996868-is-my-data-used-for-model-training (checked Oct 17, 2026), privacy.anthropic.com/en/articles/7996866-how-long-do-you-store-my-organization-s-data (checked Oct 17, 2026), anthropic.com/legal/commercial-terms (checked Oct 17, 2026), trust.anthropic.com (checked Oct 17, 2026) |
| Retention after you delete | Within 30 days (policy-violation flags: up to 2 years; scores up to 7 years) Deleted chats are not used for future training. Source: privacy.claude.com/en/articles/10023548-how-long-do-you-store-my-data (checked Oct 17, 2026) | Not tracked for business plans |
| Human review of chats | Only for feedback you send or Usage Policy enforcement Employees cannot access conversations by default. Exceptions: you share them via feedback, or Trust & Safety review is needed. Source: privacy.claude.com/en/articles/10458704-how-does-anthropic-protect-the-personal-data-of-claude-users (checked Oct 17, 2026) | Not tracked for business plans |
| Temporary / incognito chat | Yes: Incognito chats Ghost icon on a new chat (outside projects). Not saved to history or memory, not used for training, kept 30 days. Source: privacy.claude.com/en/articles/10023580-is-my-data-used-for-model-training (checked Oct 17, 2026), support.claude.com/en/articles/12260368-use-incognito-chats (checked Oct 17, 2026) | Not tracked for business plans |
| Data location / residency | US and other countries (no consumer region choice stated) Source: anthropic.com/legal/privacy (checked Oct 17, 2026) | Unknown Not stated on the Anthropic pages we checked. Pages checked: privacy.anthropic.com/en/articles/7996868-is-my-data-used-for-model-training (checked Oct 17, 2026), privacy.anthropic.com/en/articles/7996866-how-long-do-you-store-my-organization-s-data (checked Oct 17, 2026), anthropic.com/legal/commercial-terms (checked Oct 17, 2026), trust.anthropic.com (checked Oct 17, 2026) |
| SOC 2 Type II | Not checked yet Not checked in this personal-plan research pass (consumer/business compliance docs not reviewed). | Yes Trust Center scope: SOC 2 Type 2. Source: trust.anthropic.com (checked Oct 17, 2026) |
| ISO/IEC 42001 | Not checked yet Not checked in this personal-plan research pass (consumer/business compliance docs not reviewed). | Yes Trust Center: ISO/IEC 42001:2023. Source: trust.anthropic.com (checked Oct 17, 2026) |
| DPA available | Not checked yet Not checked in this personal-plan research pass (consumer/business compliance docs not reviewed). | Yes Commercial Terms incorporate the Anthropic DPA. Source: anthropic.com/legal/commercial-terms (checked Oct 17, 2026) |
| BAA / HIPAA | Not checked yet Not checked in this personal-plan research pass (consumer/business compliance docs not reviewed). | Yes HIPAA listed in Trust Center; BAA available for HIPAA-ready Enterprise plans. Source: trust.anthropic.com (checked Oct 17, 2026) |
| Zero-retention option | Not checked yet Not checked in this personal-plan research pass (consumer/business compliance docs not reviewed). | Unknown ZDR agreements exist for API; not stated for Claude for Work. Pages checked: privacy.anthropic.com/en/articles/7996868-is-my-data-used-for-model-training (checked Oct 17, 2026), privacy.anthropic.com/en/articles/7996866-how-long-do-you-store-my-organization-s-data (checked Oct 17, 2026), anthropic.com/legal/commercial-terms (checked Oct 17, 2026), trust.anthropic.com (checked Oct 17, 2026) |
Verbatim quotes from each source are on the personal and business pages.
How to turn off training on a personal Claude account
Turn off "Help improve our AI models" in Privacy settings. Step-by-step instructions →
See also: Why AI training on your data matters · How we source and verify this