AI Policy

1. You Are Interacting With an AI System

AskFredd.ai is an automated conversational AI system. When you use any of its sections (Diseases & Therapy, Drugs, Experts, News, Training & CME), you are interacting with software, not with a human being. No human operator, clinician, or physician reviews or monitors conversations in real time. A clear, non-dismissible notice is shown at the top of every chat and a mandatory acknowledgement is required before you can start.

2. Our Role Under the AI Act

For the purposes of the EU AI Act, the operator of AskFredd.ai acts as a deployer of a general-purpose generative AI system used to produce synthetic text. AskFredd.ai is intended as a limited-risk, purely informational and educational tool. It is not a medical device, and it is explicitly designed not to perform diagnosis, triage, individual risk assessment, or treatment recommendations — activities that could otherwise fall under high-risk obligations.

3. Which AI Technology We Use

Answers are generated by large language models (LLMs) accessed through the Abacus.AI platform, which routes requests to state-of-the-art foundation models (for example, models from the Anthropic Claude family). To provide up-to-date, source-based answers, the system also performs real-time web searches against a curated set of trusted scientific and institutional sources.

Because AskFredd.ai relies on third-party foundation models, the specific model version may change over time as providers release improvements. In all cases the underlying content remains machine-generated.

4. Synthetic Content — No Human Editorial Review

All responses produced by AskFredd.ai are synthetic content generated entirely by artificial intelligence and are not subject to human editorial control or verification before being shown to you. AI systems can produce information that is inaccurate, incomplete, biased, or out of date (so-called “hallucinations”). You should always verify important information against original, authoritative sources and consult a qualified professional before making any decision about your health.

5. Safety Guardrails

The AI is instructed, through its system configuration, to operate within strict boundaries. In particular, it is designed to:

• act in a purely educational and informational role;
• refuse to provide personal diagnoses, individual triage, or prescriptions;
• refuse to recommend specific dosages for individual use or changes to a person’s treatment;
• include a disclaimer with every response;
• detect potential crisis or self-harm situations and immediately redirect users to emergency and crisis-line resources.

6. Marking of AI-Generated Content

Under Article 50 of the AI Act, providers of generative AI must mark their outputs as artificially generated (for example through machine-readable watermarks or metadata). This technical obligation falls primarily on the foundation-model providers. AskFredd.ai does not remove, alter, or interfere with any such technical markings applied by those providers, and additionally labels AI interactions transparently to users through the notices described above.

7. Data You Enter in the Chat

Because AskFredd.ai deals with mental-health topics, the text you type could reveal sensitive information (special-category data under Article 9 GDPR). To protect you:

• Please do not enter your name, other people’s names, contact details, or any personal or identifying health information.
• Conversations are session-based and are not tied to an account or profile.
• Your queries are sent to the AI and web-search providers solely to generate your answer.

For full details on data handling, see our Privacy Policy.

8. Human Oversight and Responsibility

AskFredd.ai is provided as a knowledge aid and does not replace the judgement of a qualified healthcare professional. You remain responsible for how you use the information. If the tool ever behaves in a way that appears unsafe or inappropriate, please report it through the contact channels on our About page so it can be reviewed.

9. Changes to This AI Policy

We may update this AI Policy to reflect changes in the technologies we use, in our safeguards, or in applicable law. The current version will always be available on this page, with the “Last updated” date revised accordingly.

10. Related Information

Please also read our Disclaimer & Terms of Use, our Privacy Policy, and our Cookie Policy. If you or someone you know is in crisis, see our Emergency Resources page.