What the EU AI Act regulates
The AI Act is the world's first comprehensive AI law. It classifies AI systems by risk: prohibited practices (such as social scoring), high-risk systems (such as AI in recruitment), systems with transparency obligations, and minimal-risk applications. The higher the risk, the stricter the requirements.
The regulation entered into force in August 2024 and applies in stages: prohibitions since February 2025, obligations for general-purpose AI models since August 2025 – and broad application, including the transparency obligations, from 2 August 2026.
Where AI phone assistants fit in
A phone assistant that answers calls, responds to questions and books appointments is generally not a high-risk system. It does fall under the transparency obligations of Article 50: anyone deploying an AI system that interacts directly with people must ensure those people are informed about the AI – unless it is obvious from context.
On the phone, nothing is “obvious”: modern AI voices are barely distinguishable from human ones. So in practice the rule always applies – callers must learn at the start of the conversation that an AI is speaking.
What businesses should do by August 2026
- Take stock: where are voicebots, chatbots or AI assistants already used in customer contact?
- Review announcements: does every AI conversation start with a clear notice about AI use? Is the announcement hard-wired or can it be switched off?
- Hold providers accountable: the provider should contractually guarantee AI Act compliance and technically enforce the transparency features.
- Do the GDPR homework too: add the AI assistant to your DPA, records of processing and privacy policy.
- Document: record which system is used, how people are informed and who is responsible.
GDPR and AI Act apply in parallel
The AI Act does not replace the GDPR – both apply side by side. The GDPR governs how personal data (voice, name, request) may be processed; the AI Act additionally governs how the AI system itself must be designed. In practice this means EU data processing, a deletion concept and the AI announcement belong together.
Violations carry substantial penalties: depending on the breach, the AI Act provides for fines of up to €15 million or 3 percent of global annual turnover – and even more for prohibited practices.
Conclusion
For businesses that want to use AI telephony, the AI Act is not an obstacle but a clear set of rules: disclose AI use, process data in the EU, document properly. Empfango meets the transparency obligation by default – every conversation starts with a fixed announcement about AI use and recording that cannot be switched off.