EU AI Act for Telecoms: Transparency, Critical Infrastructure & High-Risk Triggers
Telecom operators use AI for network optimisation, predictive maintenance, customer service and churn prediction. Under the EU AI Act (Regulation (EU) 2024/1689), most of these are limited- or minimal-risk, with Art. 50 transparency duties for customer chatbots. Two specific areas can escalate: AI running as a safety component of critical digital infrastructure, and AI credit-checking consumers at contract sign-up.
Is it in scope?
General telecom AI — traffic forecasting, churn models, service chatbots — is not high-risk; chatbots attract Art. 50 transparency. The escalation point is Annex III(2), which lists AI used as a safety component in the management and operation of critical digital infrastructure. AI that performs a safety function in network operation can therefore be high-risk. Separately, where an operator uses AI to assess a consumer's creditworthiness before granting a post-paid contract or device finance, that scoring is high-risk under Annex III(5)(b).
Typical AI use cases
- Network traffic optimisation and management
- Predictive maintenance of network equipment
- Customer-service and support chatbots
- Churn prediction and retention targeting
- Creditworthiness checks for contracts and device finance
Risk classification
For most operators the classification is limited-risk: disclose AI chatbots and mark AI-generated content under Art. 50 (from 2 August 2026), with churn and marketing models generally minimal-risk. Escalation is specific: AI acting as a safety component in the operation of critical digital infrastructure is high-risk under Annex III(2), and consumer credit checks at sign-up are high-risk under Annex III(5)(b). Each of those features must be assessed and documented individually rather than folded into a single 'telecom AI is low-risk' assumption.
Obligations to prepare for
Your exact duties also depend on whether you build or use the AI. See obligations by operator role — provider, deployer, importer, distributor or GPAI provider.
FAQ
Is our network-optimisation AI high-risk?
It depends on function. Ordinary traffic optimisation and forecasting are typically minimal-risk. But AI used as a safety component in the management and operation of critical digital infrastructure is high-risk under Annex III(2), so safety-critical network functions need a specific assessment.
Does running a credit check on new subscribers change our tier?
Yes, for that feature. AI used to evaluate a consumer's creditworthiness or credit score is high-risk under Annex III(5)(b), even when it is only one step in a telecom sign-up flow. It triggers the full high-risk obligations.
Do our support chatbots need disclosure?
Yes. Under Art. 50 you must inform customers they are interacting with an AI system unless it is obvious from the context, applicable from 2 August 2026, and mark any AI-generated content.
Run the free 2-minute risk checker — no sign-up.