Data & AI
Safeguarding sustainable growth, efficiency, profitability, and thus also maintaining our long-term competitiveness, requires us to continually work on improving our processes. This is what drives us to keep digitalizing. We want to fully unlock the potential of our data, make innovative technologies such as AI an integral part of our daily work, and make the best possible use of the advantages of AI along our value chain. It starts with our networks: going forward, we want to make our networks increasingly autonomous, able to run with minimal human intervention. For instance, in the future, AI will enable us to individually adapt each mobile radio cell to expected workload levels. It will anticipate spikes in local utilization rates, e.g., if there’s an open-air concert or a soccer game, and increase the antenna capacity as required. Likewise, it can put certain frequencies into sleep mode during periods of expected low demand, e.g., at night or during school vacation times. Algorithms like these can help reduce electricity consumption without compromising the browsing experience. Another example is automated ticket handling. Complaints about network disruptions received from customers via the service channel are automatically converted into tickets. Since August 2024, we have been using large language models (LLMs) to create some of our tickets. These models can glean a clearer picture of the issue from the message content and add structure if needed. AI will also help resolve the tickets in the future. Deploying AI in our networks reduces outages, improves quality, and enhances our energy efficiency.
LLMs and generative AI (GenAI) have the potential to completely transform all customer interfaces. We use GenAI to improve our Frag Magenta chatbot, meaning we can also handle customer inquiries for which we have no suitable existing script, or which contain unclear or ungrammatical wording. In the future, it will further improve the automated handling of customer inquiries and reduce the volume of inbound calls to our customer advisors. In addition to evolving our chatbot, we have added a general AI assistant to Frag Magenta in our OneApp: Magenta AI was launched in the fourth quarter of 2024. Magenta AI contains both the chatbot and an AI-powered search engine that uses global knowledge from freely accessible sources. This tool opens up access to AI to the broader public. The work of our employees outside of customer interactions will likewise benefit from AI in the form of task automation, document insights, and automatic suggestions. The use of AI in IT, such as in all phases of software development, is being successfully tested: from AI-powered code generation to finished software testing, to continual error detection and resolution. Further examples include the analysis of legal texts or the use of internal chatbots to simplify access to internal data and documents. In the reporting year, we launched the AskT chatbot in collaboration with our partner Glean. Various different models are being deployed, currently those from Glean and OpenAI. AskT accesses both internal and external data sources to provide answers.
Our use of AI complies without exception with our own ethical guidelines for handling artificial intelligence.
We established the AI Competence Center (AICC) in 2023 to centrally pool the Group’s AI competencies. It delivers practical technical support for all Group entities, helping them in particular to rapidly deploy generative AI. The AICC provides advice on using AI ethically and securely, and assists with selecting suitable partners. We also play an active role in the AI ecosystem, predominantly through the Global Telco AI Alliance, which we formed in 2023 together with e&, Singtel, and SK Telecom, and to which we welcomed SoftBank in the reporting year. The Alliance’s goal is to develop new growth drivers through innovative AI models.