Skip to main content
AIRIN
PricingSign in

FlowManual policy evolution

Before/after stance changes across captured policy versions, with exact citations. If no before/after delta is available yet, AIRIN shows the latest citation-backed stance events instead.

Diffs
0
Improved
0
Worsened
0
Changed
0
No before/after stance delta is available for this filter yet. Latest citation-backed stance events are shown below.
Aug 7, 2026subprocessors / sharingmedium

data sharing

Latest stance: third party or vendor sharing

We do not sell your data. We do not use your document content for advertising. We do not share your data with third parties except as described in Section 3.
Open citation
Aug 7, 2026subprocessors / sharinghigh

data sharing

Latest stance: sale or sell

We do not sell your data. We do not use your document content for advertising. We do not share your data with third parties except as described in Section 3.
Open citation
Aug 7, 2026traininglow

model training

Latest stance: no training claim

When you run document analysis, extracted plain text from your documents is sent to Anthropic’s API. Anthropic operates under a zero-data-retention policy: prompts and responses are not stored and are not used to train models. Raw document files are never transmitted. See Anthropic’s Privacy Policy .
Open citation
Aug 7, 2026traininglow

model training

Latest stance: no training claim

When you upload a document, the PDF may be sent to Microsoft Azure Document Intelligence (United States region) to read its layout, tables, and text. Azure processes the file for that purpose only: it does not use your documents to train models, and analysis data is deleted within 24 hours. You can ask us to turn this off for your account, in which case parsing runs only on our own servers. See Microsoft’s data and privacy documentation .
Open citation
Aug 7, 2026traininglow

model training

Latest stance: no training claim

Web search inside project chat is off by default and only available when we enable it on the hosted service. When you use it, the search query (which can include text from your conversation) is sent to Tavily to retrieve results. Queries are not used to train models. Web search stays off unless we enable it for the hosted service, and it is always off when redaction is active. See Tavily’s Privacy Policy .
Open citation
Aug 7, 2026retentionlow

retention

Latest stance: deletion or time bound

Your data is retained until you delete it or request account deletion. Deleted data is removed within 30 days.
Open citation
Jul 22, 2026retentionlow

retention

Latest stance: deletion or time bound

Your data is retained until you delete it or request account deletion. Deleted data is removed within 30 days.
Open citation
Jul 22, 2026subprocessors / sharinghigh

data sharing

Latest stance: sale or sell

We do not sell your data. We do not use your document content for advertising. We do not share your data with third parties except as described in Section 3.
Open citation
Jul 22, 2026subprocessors / sharingmedium

data sharing

Latest stance: third party or vendor sharing

We do not sell your data. We do not use your document content for advertising. We do not share your data with third parties except as described in Section 3.
Open citation
Jul 22, 2026traininglow

model training

Latest stance: no training claim

When you run document analysis, extracted plain text from your documents is sent to Anthropic’s API. Anthropic operates under a zero-data-retention policy: prompts and responses are not stored and are not used to train models. Raw document files are never transmitted. See Anthropic’s Privacy Policy .
Open citation
Jul 22, 2026traininglow

model training

Latest stance: no training claim

When you upload a document, the PDF may be sent to Microsoft Azure Document Intelligence (United States region) to read its layout, tables, and text. Azure processes the file for that purpose only: it does not use your documents to train models, and analysis data is deleted within 24 hours. You can ask us to turn this off for your account, in which case parsing runs only on our own servers. See Microsoft’s data and privacy documentation .
Open citation
Jul 22, 2026traininglow

model training

Latest stance: no training claim

Web search inside project chat is off by default and only available when we enable it on the hosted service. When you use it, the search query (which can include text from your conversation) is sent to Tavily to retrieve results. Queries are not used to train models. Web search stays off unless we enable it for the hosted service, and it is always off when redaction is active. See Tavily’s Privacy Policy .
Open citation
Jul 20, 2026privacymedium

data sharing

Latest stance: third party or vendor sharing

We do not sell your data. We do not use your document content for advertising. We do not share your data with third parties except as described in Section 3.
Open citation
Jul 20, 2026retentionlow

retention

Latest stance: deletion or time bound

Your data is retained until you delete it or request account deletion. Deleted data is removed within 30 days.
Open citation
Jul 20, 2026traininglow

model training

Latest stance: no training claim

When you run document analysis, extracted plain text from your documents is sent to Anthropic’s API. Anthropic operates under a zero-data-retention policy: prompts and responses are not stored and are not used to train models. Raw document files are never transmitted. See Anthropic’s Privacy Policy .
Open citation
Jul 20, 2026privacyhigh

data sharing

Latest stance: sale or sell

We do not sell your data. We do not use your document content for advertising. We do not share your data with third parties except as described in Section 3.
Open citation
Jul 20, 2026traininglow

model training

Latest stance: no training claim

Web search inside project chat is off by default and only available when we enable it on the hosted service. When you use it, the search query (which can include text from your conversation) is sent to Tavily to retrieve results. Queries are not used to train models. Web search stays off unless we enable it for the hosted service, and it is always off when redaction is active. See Tavily’s Privacy Policy .
Open citation

Generated from live stance events. Informational only, not legal advice.

📢 POLICY UPDATES ALERT

AIRIN Brief

Built for compliance officers, legal counsel, and SaaS founders. Subscribe to the email digest — one short brief when a tracked vendor materially changes its terms, training policy, or risk rating. Prefer in-app? Watch platforms in your alerts inbox instead.