Fruitly and the EU AI Act

As an AI lab from Austria, we run Fruitly under the rules of the European AI Act – and show you how to check that yourself.

Last updated: October 2026

What it’s about

The AI Act (Regulation (EU) 2024/1689) regulates artificial intelligence in the EU by risk: the greater the possible impact on people, the stricter the rules. It applies step by step.

  1. August 1, 2024

    The regulation enters into force.

  2. February 2, 2025

    Prohibited AI practices and the duty of AI literacy apply.

  3. August 2, 2025

    Rules for general-purpose AI models apply.

  4. August 2, 2026

    Most remaining rules apply – including the transparency obligations for AI systems.

Where Fruitly fits

Fruitly is an AI assistant for general tasks: answering questions, research, writing, creating documents and websites. It isn’t meant for the high-risk areas the AI Act regulates most strictly – such as recruitment, credit scoring or decisions about access to education and public services. We don’t offer or allow practices prohibited by Art. 5, such as manipulative techniques or social scoring.

What we do

You know you’re talking to AI

Art. 50(1)

Fruitly always presents itself as an AI assistant and never pretends to be human.

AI text can be recognized

Art. 50(2)

Every answer carries an invisible watermark. It changes neither content nor quality, is always on and can only be detected with our secret key. Anyone can check with the AI Detector.

Run in Austria

GDPR

Our model runs on our own servers in Austria. Your chats never go to another AI provider.

Your data, your choice

GDPR

We only train on your chats if you explicitly allow it. What you delete is gone right away and removed for good from our abuse review after 30 days. Details in our Privacy Policy.

People decide

Principle

AI can make mistakes. Fruitly shows its sources so you can check answers, and it makes no decisions about you. Double-check important answers – especially on health, law and money.

AI literacy on our team

Art. 4

We make sure everyone working on Fruitly has sufficient knowledge of AI and its risks.

Check it yourself

Reports and questions

Did an answer seem harmful, wrong or discriminatory? Tell us in the app under Settings › Contact Support or by email. We look into every report.

Source: Regulation (EU) 2024/1689 on EUR-Lex