Anthropic's July 2026 Privacy Update: What Claude Users Should Check
Anthropic's updated Privacy Policy takes effect July 8, 2026 for Claude Free, Pro, and Max users, adding clearer language around connected apps, multi-step tasks, verification data, studies, third-party data flows, and model-improvement controls.
What changed in Anthropic's July 2026 Claude privacy update
Anthropic has published a Privacy Policy update that takes effect on July 8, 2026, and applies to Claude Free, Pro, and Max consumer accounts. The update does not apply to Claude Team, Claude Enterprise, the Claude Developer Platform, or other services governed by Anthropic's Commercial Terms or separate agreements. The practical change is clearer disclosure around Claude acting across apps, handling multi-step tasks, collecting verification data in certain cases, and using account or study data under defined privacy settings.
Key takeaways
- The update is limited to consumer accounts: Claude Free, Pro, and Max.
- Anthropic now explains more about data flows when Claude works with third-party apps, connectors, plugins, webhooks, external APIs, or other connected services.
- The new policy describes verification data, including age or identity checks that may involve government ID, selfie or video data, and verification results.
- Study participation data is called out for users who join Anthropic studies, surveys, interviews, or related research.
- Anthropic says Claude remains ad-free, it does not sell user data, and users can control whether conversations are used to improve Anthropic's models, subject to the policy's safety and reporting exceptions.
Practical LinkLoot angle
This update matters because Claude is becoming less like a static chatbot and more like an agent that can act across services. Once a user connects apps or asks Claude to complete work outside the chat window, privacy is no longer only about the prompt text. It also includes files, third-party content, actions Claude takes, data Claude retrieves, and the permissions attached to each integration.
For everyday users, the best response is not panic. It is an account-settings audit:
| Area | What to check | Why it matters | Source |
|---|---|---|---|
| Consumer account scope | Whether you use Free, Pro, or Max versus Team, Enterprise, Platform, or API | The July 2026 update is aimed at consumer accounts, not commercial services under separate terms | Anthropic Privacy Center |
| Connected apps | Which third-party services Claude can access | Claude may send instructions, inputs, and outputs to connected services and retrieve data from them | Anthropic Privacy Policy |
| Verification prompts | Whether a feature asks for age or identity verification | The policy describes ID document data, selfie or video data, facial geometry templates, and verification results in certain cases | Anthropic Privacy Policy, TechCrunch |
| Model-improvement settings | Whether your chats and coding sessions are enabled for model improvement | Anthropic says users can control this setting, while the policy also keeps safety review and explicit-reporting exceptions | Anthropic Privacy Policy |
The workflow is simple: open Claude's privacy settings, review model-improvement preferences, disconnect integrations you do not actively use, and treat every connected service as a separate privacy boundary. If Claude can read, write, send, or retrieve data through a connected app, the app's own privacy policy and permissions matter too.
What to verify before you act
First, verify the exact account type. A personal Claude Pro account and a company-managed Claude Team or Enterprise account are not governed the same way. Second, check whether Claude has ongoing access to any third-party service, not just whether you used a connector once. Third, if you see an age or identity verification prompt, confirm what capability triggered it and which provider handles the check before submitting documents.
The update also makes model-training controls worth checking again. Anthropic's current policy says it may use Inputs and Outputs to train and improve AI models unless the user opts out through account settings, with separate handling for flagged safety review material and explicitly reported feedback. That is a settings decision, not something to leave on autopilot.
The Privacy Policy page says it is effective July 8, 2026, and was published on June 8, 2026.
For a broader checklist on agent workflows and connected tools, pair this update with LinkLoot's AI workflow automation guide. The useful question is not only what the model can do, but which data and permissions each workflow requires.
Source check
Anthropic's Privacy Center confirms the July 8, 2026 effective date, the consumer-account scope, and the summarized changes around multi-step tasks, connected apps, verification data, study participation, and data practices. Anthropic's full Privacy Policy confirms the published date, the categories of data listed in the policy, the third-party service data-flow language, and the model-improvement opt-out framing. TechCrunch independently reported on the identity-verification angle and tied it to the new version of Anthropic's policy, adding outside context on why users may notice verification prompts.