Does Instinct train on your data?
Instinct trains on what its assistant sees — your emails, messages, screen and audio — by default. There is a switch at app.instinct.com/settings; safety-flagged material is trained on regardless.
A self-serve opt-out exists
"to develop and improve our products and technologies, including to evaluate, fine-tune, and train the AI models that power our products and technology. If you do not want us to use your information to train AI models, you can opt out by visiting app.instinct.com/settings"
- 21 Sept 2026
- All accounts
- global
- Fetched from the company
- 72F35D63…1D7A
How we read it: What the assistant sees is what is trained on: the policy says that "may include the content of messages, emails, or other private communications", plus audio and voice if enabled. The switch is free, but "Even if you opt out, we may still use your information for AI model training when that information is flagged for safety review", the opt-out "applies on a go-forward basis", and models "previously trained, fine-tuned or improved on your information" keep it. Two carve-outs: information from Google Workspace APIs is never trained on, under Google's Limited Use rule, and the terms exempt material placed in the "Vault". Scored above Anthropic's 50 for the breadth in scope.
opt out, opt-out, settings, object, withdraw, default, delete, Vault
What the document does offer
- in the privacy policy, the same sentence as the training use: "If you do not want us to use your information to train AI models, you can opt out by visiting app.instinct.com/settings"
- in the privacy policy, immediately after: "Even if you opt out, we may still use your information for AI model training when that information is flagged for safety review"; "This opt out applies on a go-forward basis"; "We may still use AI models previously trained, fine-tuned or improved on your information prior to your opting out"
- in the privacy policy: "We do not use information received from Google Workspace APIs to evaluate, fine-tune, train, or improve AI models"
- in the terms (instinct.com/terms, captured 2026-09-08): the same opt-out and exception for "Materials", plus "for Materials that you add to our feature of the Service known as the 'Vault' ... we will not use such Materials to train AI models"
- "default": 0 matches in either document
Why none of it applies: The control is stated in the same sentence as the training use and given a location, app.instinct.com/settings, though not a named switch, so this is Airbnb's pattern rather than a named toggle. What it does not do is reach back: the opt-out is "go-forward", models already trained keep your data, and safety-flagged material is trained on regardless. "Default" appears nowhere; training is simply what happens unless you visit the page. The terms repeat the same opt-out and exception word for word for "Materials".
What each plan can refuse
The same questions, asked of every plan we could identify. Every cell comes from a quoted clause or a documented search - see the evidence above.
- Self-serve toggle
- Not stated
- 54 / 100
- Not stated
- Unknown
How to opt out of Instinct AI training
Plan by plan, from what Instinct's own policy says about refusing. Where it publishes the steps, they're listed here with the limits they carry. Where there's nothing to switch off, or nothing graded, this says so instead of inventing a menu path.
All accounts
Yes. Instinct has a self-serve setting.
- Go to app.instinct.com/settings, the page the policy names. It does not name the control on that page.
- Material you add to the Vault is not trained on at all, whatever the switch says.
What this policy has done since we started watching
No change detected. We've held a copy of Instinct's documents since 8 Sept 2026 and re-read them since, across 4 captures of 2 documents. The text hasn't moved in a way that touches this verdict. That's a fact about the documents, and it's only knowable by having kept the earlier copies.
Vendor response
We wrote to Instinct on 15 Sept 2026, quoting the clauses above and inviting a correction. Nothing has come back. That's published because a company declining to comment on its own policy is something a reader should know, and because the invitation should be on the record whether or not it's taken up. Any reply goes here, word for word.
Report a wrong clause