AI training and your data: our commitment

How Fyxer uses your data to train its AI models, the safeguards we apply, and how to opt out.

We're often asked whether Fyxer uses customer data to train AI models. This article explains exactly what we do and don't do, and the choices you have.

The short answer

Fyxer trains its own AI models to make features like drafting, inbox sorting, and transcription better. These models are private to Fyxer. We never let other AI providers train their own models on your data, and you can opt out of Fyxer's training at any time (see below).

Your training setting

The Allow Fyxer to use my data to train and improve its AI models setting is on by default. To opt out:

1. Go to Dashboard → Manage your account

2. Uncheck Allow Fyxer to use my data to train and improve its AI models

Admins can opt out the whole organization by selecting: Opt everyone in [organization name] out of model training. You can also ask us not to use your data for training by emailing privacy@fyxer.com.

If your browser sends a Global Privacy Control signal when you sign up, we turn off this setting from the start.

What opting out does

• Your data won't be used in any future training

• Data already used can't be removed from models that have been built, but it won't be used again. Before use, data was pseudonymized and sensitive information removed

• Your Fyxer features keep working as before, including drafts written in your voice

Personalization and training are different

Fyxer learns your writing style and preferences so it can draft in your voice. This happens each time Fyxer prepares something for you, and it doesn't change any AI model. It's part of how Fyxer works, so it stays on if you opt out of training.

Safeguards before training

• We replace names and email addresses with codes

• We remove sensitive information

• We don't use any content that hasn't been through these steps

• The models we build are private to Fyxer. They're never shared with other providers or added to their general AI models

What Fyxer never does

Fyxer doesn’t:

• Sell your data

• Use it to build advertising profiles

• Make it public

• Use it to make automated decisions that have a legal or similarly serious effect on you

• Let third-party AI providers train their own models on it

When people at Fyxer can see your content

Fyxer is designed to run without people reading your content.

In limited cases, our staff and people contracted to us may review specific content, for example to check the quality of Fyxer, investigate a fault, a security incident, or suspected misuse, or respond to a request you've made.

When this happens, access is approved case by case and only for a limited time. Each request must be linked to a logged change or incident before it's approved, and access is read-only. Everyone with this access is covered by strict confidentiality obligations.

Usage and performance data

Fyxer also collects information about how the service is used and performs, such as feature usage, error rates, and system health. We use this to keep Fyxer reliable. We only share it outside Fyxer in an aggregated, anonymized form that can't identify you or your organization.

Enterprise and compliance

If your organization needs written confirmation of how Fyxer handles your data, including a copy of our Data Processing Agreement, contact support@fyxer.com. For privacy questions, contact privacy@fyxer.com.

How this differs from team visibility

This article covers how Fyxer uses your data for AI training. If you want to know who inside your organization can see your data, see: Who can see your data in a team.

Still have questions?

Head to the chat in the bottom right of your Dashboard.