1. Who is responsible
AI.info is an independent publication about artificial intelligence, created and edited by Gabriele Masetti, who is the data controller for what is described here. For anything about your data, write to privacy@ai.info or use the contact form.
2. What we collect, and when
Reading the site: nothing about you. There is no account for readers, no cookie, and no analytics. The pages, the feeds and the calendar can be read without telling us anything.
The contact form. Your name, email address, the subject and the message, kept so we can answer you. Messages are deleted after twelve months unless the correspondence is still open.
Email lists, only if you subscribe. The weekly letter, the job alerts and the research digest each keep your email address, the filters you chose, the time of your request, the IP address it came from (the record of your consent), and the confirmation and unsubscribe tokens. Nothing is sent until you confirm from the email we send you. Every message carries an unsubscribe link, and none carries a tracking pixel or a tracked link: we do not know whether you opened or clicked anything. When you unsubscribe, your address is kept only as the record that you did, so you are not written to again; ask and it is deleted outright.
News on Telegram, only if you start the bot. If you write /start to our Telegram bot, we keep your Telegram chat id (a number), the language of the messages, the topics you chose, whether you want each story or one digest a day and at what hour, when you subscribed and when the bot last wrote to you. Not your name, your username or your phone number. Telegram carries the messages under its own privacy policy. Send /stop and the record is deleted at once; if you block the bot, it is deleted at the next message it tries to send you.
Follow, only if you follow something. When you follow a company, a topic or a desk, we keep what you follow and when, the language of the messages, and the way you chose: your email address (nothing is sent until you confirm from the email we send you, every letter carries a one-click unsubscribe link, and none carries a tracking pixel or a tracked link) or your Telegram chat id (a number, once you press Start in our bot). To send each story only once, and no more than a few a day, we note which stories went to you and when, for sixty days. Not your name, your username or your IP address. There is no password: a link we send you opens the page where you manage it, and your browser remembers that link until you tell it to forget it. “Stop everything and delete” on that page, or /stop in the bot, deletes the record at once.
Search in words. On the jobs and events pages you can type a sentence instead of setting filters. The sentence, and nothing else, is sent to a language-model provider (DeepSeek) to be turned into filters. We keep the sentence and the filters it became for ninety days, with no identifier of any kind, to see what people look for and do not find. To prevent abuse, requests are counted per IP address in hour-long windows; the counter serves nothing else.
Submissions. When you submit a tool, a job, an event or a creator profile, we keep what you entered, including the contact you gave, for as long as the listing exists or the review is open.
Signing an appeal. To sign an appeal published here, you write to the address the appeal gives. We receive your email address and what you write: your name, your country and, if you give them, your role, affiliation, a link and a short comment. The email address is used only to verify the signature; it is never published or shared, and it is deleted when the appeal closes. If you agree to be listed, the name, country, role, affiliation and link of a verified signature are published among the signatories when the appeal is delivered to the recipients it names, delivered with it, and kept online for as long as the appeal is. Some reasons for signing, if respectful and on topic, may be quoted, at the editors’ choice, on the appeal’s page and in the letter to its recipients. Write to the same address, and your signature is deleted; a list already delivered cannot be recalled.
In your browser, not with us. The theme, the accessibility settings, saved items and reading progress are kept in your browser's storage. A randomly generated identifier is kept there too and travels with a vote on a tool or a paper, so the vote is counted once; it names a browser, not a person, and is never joined to an email address or a name.
Server logs. Our hosting (Netlify) and our database (Supabase) record IP addresses, browser types and pages requested, as every host does, for security and operation. We do not use these logs to profile anyone, and they are kept for the periods those providers state.
3. What we do not do
No advertising network, no Google Analytics, no Facebook Pixel, no fingerprinting, no cookie of any kind. The site's fonts are served from our own domain, so opening a page sends nothing to Google. Our own images are served through Netlify, our host. The one exception is a poster that belongs to an event's organizer and is still hosted on their site (lu.ma, Meetup and the like): when a page shows it, your browser fetches it from there, under that site's own policy.
4. What we use it for
- To answer the message you sent.
- To send you exactly what you subscribed to, until you unsubscribe.
- To turn a sentence into filters, once, when you ask for it.
- To publish the signature you sent to an appeal, and deliver it with the appeal, once it is checked.
- To keep the site running and secure, and to investigate an incident.
- To meet a legal obligation.
We never sell, rent or trade personal data, and never use it for advertising. The legal bases are your consent (the lists, the search, a signature), the answer you asked for (the contact form, the submissions) and our legitimate interest in running a secure site (the logs).
5. Who processes it for us
Netlify hosts the site and serves its images; Supabase holds the database, the storage and the functions that read the sources; Resend delivers the emails you subscribed to; DeepSeek receives the sentences typed into the search in words. Each acts on our instructions under a data processing agreement. Some of them are established in the United States: where personal data leaves the European Economic Area, it does so under the safeguards those agreements provide, the European Commission's standard contractual clauses. The models that help write and verify the site's own content never receive reader data.
6. Your rights
Under the General Data Protection Regulation and the laws that apply to you, you can:
- see the personal data we hold about you, and receive a copy;
- have it corrected, or deleted;
- withdraw a consent at any time, with no effect on what came before;
- object to, or restrict, a processing;
- complain to your supervisory authority (in Italy, the Garante per la protezione dei dati personali).
Write to privacy@ai.info; we answer within a month.
7. Children
AI.info is not addressed to anyone under sixteen, and we do not knowingly keep data about a child. If you believe a child has sent us something, tell us and it will be deleted at once.
8. Changes
When the site changes what it keeps, this page changes with it, and the date at the top says when. A change that affects you materially — a new use, a new processor — is announced on the page itself, not slipped in.