How we care for what we are trusted with.

Everyone sees what corresponds to their role and institution: a student sees their own process, a teacher their classes, a family the student they are linked to.

Within each institution, positions are scoped by level, class or section. No one sees more than they should.

People under 18 reach Lira AI through their institution, with the authorization and scope the institution defines.

The individual pilot is only for people aged 18 or over. We do not ask for a date of birth: confirming you are an adult is enough.

Messages and files shared with the assistant are processed with Google's Gemini API to generate answers. Under Google's commercial terms, that content is not used to train their models.

Lira AI does not make high-impact decisions about students. Signals per class flag tension; the decision always belongs to a person.

When you apply we store your answers, your email and how you reached the site, to review the application and build a varied cohort.

An AI helps read the answers, but without your name, email or origin, and it does not infer personality, intelligence or socioeconomic status. It does not approve or decline: a person on the team decides.

Applying does not create an account or guarantee access. We keep applications for up to 12 months after the pilot closes, then delete or anonymize them.

On the website we measure visits and form steps, without name, email or phone number.

During the pilot we measure feature usage and feedback. Never the content of chats, files, audio, camera, keystrokes, screen recordings or precise location.

You can request access, correction, export or deletion of your data — including your application — by writing to us. Detailed activity is kept for up to 180 days; after that only anonymous aggregates remain.

Personal data is processed under Argentine Law 25.326 on Personal Data Protection.