Does Grammarly train on your data?
Grammarly · Writing assistant · official site ↗
Depends on your planZDR / DPA on enterpriseOpt-out available
Does Grammarly train its models on your data?
It depends — Grammarly trains on consumer/free-tier data by default but excludes business, enterprise and API tiers.
Grammarly's "Product Improvement and Training" is on by default for individual and team accounts (you can opt out), but Grammarly never lets third-party LLM providers train on your content and does not sell it.
Can you opt out?
Individuals (Free, Premium, single-user Pro): turn off Product Improvement and Training at account.grammarly.com/security/privacy. For Team/Business accounts it is initially on and an admin can disable it org-wide. When off, your content is not used for training or product improvement.
Zero retention / DPA
Grammarly does not allow LLM service providers to train on user content and does not sell content. Enterprise agreements add a DPA and contractual data-use restrictions; confirm specific retention terms with Grammarly for your contract. DPA ↗
What the listicles get wrong
Two layers: Grammarly's own product-improvement training is opt-out (default on) for individuals and teams; third-party LLM training is never permitted. Confirm enterprise retention specifics in-contract.
Verdict by plan tier
Free / Premium / Pro (individual)Trains unless you opt outProduct Improvement and Training is on by default; toggle it off in privacy settings to stop your content being used.
Team / BusinessTrains unless you opt outOn initially; an admin can opt the whole organization out.
Third-party LLM providersNo trainingGrammarly never permits external LLM providers to train on your content, on any plan.
Get notified when this changes
We track Grammarly's data-training and retention policy. Leave your email and we'll send one note if it changes.
Frequently asked questions
Does Grammarly train its AI models on my data?
It depends — Grammarly trains on consumer/free-tier data by default but excludes business, enterprise and API tiers. Grammarly's "Product Improvement and Training" is on by default for individual and team accounts (you can opt out), but Grammarly never lets third-party LLM providers train on your content and does not sell it.
Can I opt out of Grammarly training on my data?
Individuals (Free, Premium, single-user Pro): turn off Product Improvement and Training at account.grammarly.com/security/privacy. For Team/Business accounts it is initially on and an admin can disable it org-wide. When off, your content is not used for training or product improvement.
Does Grammarly offer zero data retention (ZDR) or a DPA?
Grammarly does not allow LLM service providers to train on user content and does not sell content. Enterprise agreements add a DPA and contractual data-use restrictions; confirm specific retention terms with Grammarly for your contract.
Is Grammarly safe to use with confidential or proprietary data?
It depends on your plan tier. Free / Premium / Pro (individual): Product Improvement and Training is on by default; toggle it off in privacy settings to stop your content being used. Team / Business: On initially; an admin can opt the whole organization out. Third-party LLM providers: Grammarly never permits external LLM providers to train on your content, on any plan. Always confirm current terms with Grammarly before sending confidential data — this is cited public information, not legal advice.
Sources
https://support.grammarly.com/hc/en-us/articles/25555503115277-Product-Improvement-and-Training-Control
https://www.grammarly.com/privacy-policy
https://www.grammarly.com/trust
This page is cited public information, not legal or compliance advice. Whether Grammarly trains on your data, and any zero-retention or DPA option, can depend on your specific plan, region and contract. Always confirm current terms with Grammarly before sending confidential or proprietary data.