Loading

wait a moment

Grammarly Will Use Your Data to Train Its AI Unless You Buy 500 Accounts

Companies may want to rethink how they use Grammarly, and what they use it for, with the revelation that you can only opt-out of AI training by buying 500 accounts.

Grammarly is a popular writing and grammar tool, with both web and mobile clients. The company has been increasingly rolling out AI-driven writing tools, with convincing advertisements demonstrating individuals using the platform to compose emails with just the right tone.

Unfortunately, as with all AI models, Grammarly’s AI needs data to be trained, and the company is getting that data from the content users are submitting to be proofed.

Karolina Szczur, Product Designer and Co-Founder of Calibre, reached out to the company to see if there was an option to opt-out of the training, and the company’s response raises serious questions:

We currently offer a feature that permits customers to opt out of this use for Grammarly Business teams of 500 users or more.

To be clear, the only way to opt out of having one’s data be used for AI training is to purchase accounts for 500 users or more.

In describing how it collects data, Grammarly says it will “sample snippets of text at random, disassociate them from a user’s account, and then use a variety of different methods to strip the text of identifying information….In other words, we don’t store any text in a manner than can be associated with your account or used to identify you or anyone else.”

Unfortunately, as data privacy experts will tell you, there is almost no way to completely anonymize data with 100% certainty. What’s more, while Grammarly says it endeavors to protect user identities, what about proprietary data? What measures are taken to protect trade secrets, or privileged communication?

Grammarly’s use of AI and its reticence to provide a reasonable opt-out is a concerning revelation, one that will no doubt have many companies canceling their accounts, much as Calibre did.