from infrastructure and operations to AI platform and subscription, but overall value increases substantially. Aprecomm: AI is increasingly helping Telecom and Media providers reduce the real cost of service delivery— though the savings come from smarter operations rather than simple cost-cutting. By automating repetitive tasks, predicting faults before they occur, and optimising network performance, AI improves
efficiency at every level of operations. Virtual assistants and chatbots reduce call centre volumes by handling routine customer inquiries, while predictive maintenance and network automation minimise downtime and the need for manual intervention.
The impact is measurable: shorter call durations, fewer truck rolls, faster issue resolution, and lower churn all translate directly into reduced operational expenditure. At the same time, AI enables greater scalability— allowing providers to serve growing customer bases without proportionally increasing their workforce or infrastructure costs. Our customer and service provider, ACT Fibernet, experienced a 25 % decrease in customer complaints after implementing our AI-based monitoring solution.
However, AI should be viewed as a cost transformer, not merely a cost cutter. Implementation requires upfront investment in data quality, integration, and governance. The real benefit lies in long-term efficiency, improved service reliability, and enhanced customer satisfaction.
In short, AI applications do reduce the overall cost of providing telecom and media services, but their greatest value is in enabling a more intelligent, predictive, and customerfocused operating model that sustains profitability and competitiveness over time. Bitmovin: AI does help reduce total cost of ownership, but not always through direct savings. The biggest advantage lies in how it enables teams to use their existing budgets and resources more efficiently. By automating diagnostics and accelerating fault detection,
AI reduces repetitive manual testing and shortens mean time to resolution. These efficiencies allow engineering and operations teams to reallocate time and budget toward higher value initiatives such as workflow optimisation, feature development, and innovation. In this way, AI improves productivity and helps organisations scale their testing and monitoring capabilities without increasing costs.
Bridge: On the surface, perhaps- but only by taking a dangerous shortcut. True cost reduction cannot come from laying off seasoned monitoring professionals or underinvesting in emerging talent, on the assumption that AI can fill those roles indefinitely. That’ s a gamble with the operational heartbeat of a broadcast chain. AI can support expertise, streamline data analysis, and increase efficiency, but it cannot replace decades of accumulated human understanding. We see technology as a tool for empowerment, not substitution. The cost of losing that balance would be far higher than any short-term saving. Interra: The impact of AI applications on the real cost of providing T & M services depends on the type of AI solution used and the specific operational requirements. AI can indeed lower costs in areas where automation and data-driven efficiency are achievable. Tasks such as content tagging, indexing, anomaly detection, and content moderation of live or recorded streams can be handled effectively by machine learning models, reducing reliance on manual teams and improving both speed and consistency. Similarly, AI-driven analytics enable faster and deeper insights for large-scale, multi-source data operations such as OTT audience measurement, media diagnostics, and content recommendations. Leader: AI can certainly help reduce operational costs by shortening fault-finding time, automating routine analysis, and improving predictive maintenance. But the primary benefit lies in improved quality and confidence. For broadcast and production engineers, the cost of a missed fault or on-air
error far outweighs any savings from lower testing thresholds. AI augments human expertise- it doesn’ t replace it. The real value comes from enabling smaller engineering teams to maintain higher production standards with greater consistency. Telestream: AI by itself doesn’ t reduce the cost of T & M, much like the transition to cloud computing doesn’ t reduce production costs. It’ s the aggregate benefits you achieve from the technology that will reduce overall costs. An analogy to the cost savings can be gleaned from the software development process. It’ s well accepted that finding a software defect after a product is launched is significantly more costly than if the defect is found during the development and testing lifecycle. To accomplish early fault finding, in the short term, your development cycle slows down with the addition of code reviews and unit testing, costs increase for testing infrastructure and personnel to execute or automate tests. However, after launch, problems are reduced, and customers are happier and more likely to make future purchases. VeEX: The impact of AI on T & M costs depends on the AI agents and the specific tasks they are designed to support. AI does not replace all T & M functions, because one AI agent cannot do everything. This is why the industry is moving toward Agentic AI, where multiple specialised AI agents work together in sync with network equipment and dedicated T & M tools to handle tasks such as troubleshooting, maintenance, proactive monitoring, outage prevention, and service assurance. AI on servers and network equipment alone does not reduce T & M costs, because many specialised T & M functions cannot simply be replaced or embedded into network equipment. But when Agentic AI is used together with T & M devices, it can improve efficiency and lead to lower operational costs over time. VIAVI: Looking at the full TCO of the task( e. g. doing an integration between a RU and a DU, even from single vendors, and ensuring that you get the best performance out of this) is a very expensive task today and over the past generations. This involves the lab builds, the manpower for QA, Integration, SW development etc and to a relatively small portion the T & M equipment costs. Changing this to an AI driven model( using for example the rApp / xApp model for tuning the interworking between the RU and DU) can substantially lower the cost of the tuning.
Does fragmentation of service provision still present challenges? Are standards emerging that can help with this? Accedo: Yes, fragmentation remains a major challenge, but standards are emerging. Service provision fragmentation is very
EUROMEDIA 11