Fronset

Community Notice

What Fronset's social media accounts do with public posts they find and reply to, and how to make them stop.

KAPUA Labs LLC, doing business as Fronset — community accounts

Version 2026-10-28

Fronset’s accounts on X, Bluesky, Mastodon and Reddit take part in public conversations about the cost and quality of large language models. This page explains what that means for you if one of our accounts finds or replies to one of your public posts. It supplements our Privacy Policy.

To make us stop: email privacy@fronset.ai with your handle and the platform, or reply “stop” to one of our replies. When we see your request, we stop replying to you and mark your account so that we do not engage with you again on that platform.

1. Our replies are drafted with AI

Replies from our community accounts are drafted with the help of AI language models, using a system we built and run. A person sets the rules the system follows, such as the topics, the tone and what it may and may not say. Whether a person checks each reply depends on the platform:

  • X: a person reads each reply and posts it by hand.
  • Bluesky, Mastodon and Reddit: replies are posted automatically. A person does not read each one before it is posted, but automatic checks run on every reply, and a person looks over our accounts’ activity.

Our system also decides automatically which posts to reply to. That decision has no legal or similarly significant effect on you.

If one of our replies is wrong or unwelcome, tell us at privacy@fronset.ai and we will remove it.

2. Who we are

The controller is KAPUA Labs LLC, doing business as Fronset, 30 N Gould St Ste R, Sheridan, WY 82801, United States.

Privacy contact: privacy@fronset.ai.

3. What we collect, and from where

Our system searches public posts on X, Bluesky, Mastodon and Reddit for topics we follow, such as the cost, quality and evaluation of AI language models. Public X posts reach us through twitterapi.io, a third-party service that supplies public X posts. When it finds a post, we keep:

  • your public handle and display name;
  • the text and link of your post and, on Reddit, of replies to our replies that our system picks up;
  • what our system decided about your post (whether to reply, and why), and the reply we posted;
  • how many times we have replied to you and when, so that we do not reply to the same person too often, a short status for our relationship with you, and any notes we add;
  • the public engagement counts on our own reply (for example likes and reposts), and how often the link in our reply was clicked.

Some of our replies include a short link. The short-link service, TinyURL, records each click on it, including the clicking device’s IP address, approximate country, browser, device type, language and the page the click came from. We receive these records and use them to count how many people clicked each reply. They describe whoever clicked the link, who may not be the person we replied to.

We do not sell this information or share it with advertisers.

We use this information to take part in public conversations about the cost and quality of AI language models, to offer useful answers, and to make Fronset known.

Our legal basis is our legitimate interest in doing that (Article 6(1)(f) of the GDPR and of the UK GDPR). We have weighed that interest against yours:

  • we only read posts that are already public, on the topics we follow;
  • we limit how often we reply to the same person, and how many replies our accounts post automatically each day;
  • you can make us stop at any time, without giving a reason.

5. Who receives it

  • The platform where we reply. Our replies are public.
  • twitterapi.io, which supplies public X posts to us.
  • Model providers whose AI models help us decide whether to reply and draft our replies. They receive your handle and the text of your post. They are among the model providers listed in our Subprocessor Register. Some are established outside the EU/EEA, the UK and the United States, including in the People’s Republic of China; the register names them.
  • TinyURL, which runs our short links and records clicks on them.
  • Microsoft Azure, which hosts our systems in the United States, and Sentry, which may receive error reports from our systems.

6. Transfers outside the EU/EEA and the UK

We are based in the United States, and our systems run there. The information described here is therefore processed outside the EU/EEA and the UK, in the United States, and by some model providers in other countries, including China. These countries may not protect personal data to the same standard. The information we start from is already public. Write to privacy@fronset.ai if you want to know more about a particular transfer.

7. How long we keep it

We have not yet set fixed retention periods for these records, and we do not delete them automatically. Until we do, we keep the record of our interaction with your account (your handle, the posts, our replies, the click records and the counts above). If you ask us to delete it, we will.

If you asked us to stop, we keep your handle and the platform, marked so that we do not engage with you again, for as long as we need to honour your request.

8. Your rights

  • You can object at any time to our replies and to our use of your posts to promote Fronset. We then stop, without asking why.
  • You can ask for a copy of what we hold about you, and for its correction or deletion, or for its use to be restricted. You can also object to any other use we make of it.

Write to privacy@fronset.ai, with your handle and the platform. We answer within one month.

9. Complaints

You can complain to us first, at privacy@fronset.ai with “Complaint” in the subject. You can also complain to a supervisory authority: in the EU/EEA, to the supervisory authority in the country where you live, where you work, or where you think the infringement happened; in the United Kingdom, to the Information Commission (ico.org.uk).

Change log

  • 2026-10-28 — First version.