Since the European Union adopted the final version of the AI Act in March 2024, we are regularly asked: Does FirstRing fall under the category of “high-risk AI”?
A valid question—especially in crisis situations, where due diligence and trust are crucial. In this blog post, we clearly explain why FirstRing does not fall into the high-risk category, and what we do to ensure that our use of AI is safe, ethical, and transparent.
What exactly does the AI Act say?
The AI Act categorizes AI systems based on risk. Systems that autonomously make decisions concerning people's physical safety, health, or rights (such as medical diagnoses or social credit scores) fall under high-risk AI. These systems must meet stringent requirements, including risk management, logging, human oversight, and transparency.
Why FirstRing is not classified as high-risk
Based on legal advice, the following applies:
- FirstRing provides support, but does not make decisions. The tool generates communication suggestions for crisis teams, but does not make decisions on its own. The end user always decides what information is released to the public.
- FirstRing is not a safety-critical system. It does not affect any physical processes or systems (such as traffic or medical equipment) and is therefore not covered by the “safety-related use cases” under the AI Act.
- FirstRing always requires human input. The tool is designed to help people think, write, and make decisions—not to replace them.
How we DO comply with important AI principles
Even though FirstRing is not a high-risk system, we take the principles of the AI Act seriously. That is why:
- We always state that AI has been used in generated texts (transparency).
- The end-user remains responsible for what is published (human oversight).
- We offer training on ethical and responsible AI use.
- Users can choose models hosted within the EU (such as Mistral).
Want to know more?
Would you like to receive legal advice from us or learn how your organization can use FirstRing responsibly? Feel free to contact us for a demo, documentation, or a consultation.