1. Value capture problem: Hyperscalers captured the lion's share of economic value while operators financed the networks. AI infrastructure — data centres, edge computing, network intelligence — offers a breakthrough opportunity, but requires fundamentally different strategy and governance.
2. Cost transformation pressure: AI is a force multiplier for cost measures in network operations and customer service. But capturing this potential requires enterprise-wide AI transformation, not isolated pilots.
3. NIS2 and EU AI Act: Telecom is classified as an essential service under NIS2. AI chatbots directed at consumers require transparency under the EU AI Act. Network-critical AI systems potentially fall under high-risk classification.
4. Talent and literacy: The transition from network engineers to AI engineers requires massive upskilling. Implementing EU AI Act Article 4 literacy obligations at scale exceeds the internal capacity of most operators.