Slides
Page 2
AI IS AN ENTERPRISE REDESIGN AI is not just another tool,
Page 3
“Fix the basics”: IP everywhere: 100% IP accesses Service transformation: customer and digital first Convergence Push: Magenta EINS AI Guidelines: build principles for the responsible use of AI (2018) AI-native by design Network runs itself: autonomous operations as the default AI-driven Service: agents resolve the majority of interactions Connectivity for agents: monetizable, low-latency Scaling new AI revenue pools Human-led agent teams Cycle 1: Leading EU telco Step up in software and orchestration capabilities: From buyer to builder: Access 4.0, one API layer for Europe, first O-RAN Scaling digital experts and enabling an agile organization Proving transatlantic scale with OneApp Cycle 2: Leading Digital T elco Cycle 3: Data-informed, AI-enabled, digital-first DT built the digital foundations early – AI is the next step 3 Infuse AI into all processes and areas of DT AI for everyone: AI CC & AI Factory; all employees enabled AI in production: 500+ internal projects live Sovereign AI: Industrial AI Cloud with Nvidia – Europe’s largest New revenue pools with AI-native products and services Next: Leading AI T elco CMD 2015 CMD 2021 CMD 2024
Page 4
From bits & bytes to a new monetizable intelligence: powering Physical AI with low latency services (6G) & sovereign AI Clouds From selling ICT to powering enterprise AI: with sovereign, scalable and industry- specific AI solutions Network From people in queues to agents running the workflows: democratizing coding, AI- assisted decisions and automated processes Customer Enterprise AI is rewriting telco economics across three dimensions 4 ACCESS EDGE CORE 1 2 3 Enabler: a new AI operating model From functional hand-offs to human-led agent teams: redesigning, building and scaling AI E2E AI-native teams: Tiger teams, AI restructuring offices, business ownership One Group approach: common design, reusable data products, local integration AI by default: AI-native development, reusable harnesses & strong evals Value accountability: business-owned adoption along audacious targets Run cost Resilience New revenues Cost-to-serve NPS Retention AI Revenues Productivity Cycle times G&A
Page 5
AI-native products as new revenue pools 5 Industrial / Enterprise AI AI enterprise solutions across the whole stack AI-ready connectivity More intelligent, low latency, tokenizable networks for the physical AI era ACCESS EDGE CORE €1 bn + revenue potential until 20301 Consumer AI The trusted AI companion for everyday life. Smart. Secure. Device- & App agnostic 1 DT ex US
Page 6
People • >250k AI enablements since 2023 • 24.8% share of digital experts • AI toolsets deployed From trusted connectivtity to trusted AI: Our right to play in AI 6 Trust & Control • #1 Telco brand in the world • Reliable partner for government services • Europe’s largest SoC DT plays where trust and control matter most: combining sovereignty and security with access to leading-edge technology Our Game Frontier models: We don’t build the next big LLM – we partner with the best Hyperscale data centers: We don’t compete on scale – we focus where specific customer needs differentiate us Not our Game Our right to play Context & Scale • ~200 mn customers • 100 bn calls on our network • 100 mn contact center interactions p.a. Infra assets • Leading network infra- structure across EU & US • Sovereign compute in EU with AIIC live • E2E sovereign AI stack Ecosystem access • Partnerships with all leading frontier models • >€300 mn investment in leading AI companies (n8n, BFL, Lovable, Nscale, Decagon, …) From trusted connectivity to trusted AI Unmatched customer access and insights on both side of the Atlantic A sovereign AI infrastructure – already built, already live. Direct access (and insights) to the heart of the AI ecosystem Our people are AI-ready – all set and motivated to embrace the AI journey
Page 7
Sustainability, Trust and Resilience are key differentiators AI Capability Scale Value Scale AI with efficient energy, compute and resources Enable safe and reliable AI use for customers and business Ensure continuity when systems, sites and supply are under stress Efficient technology Trust & security Resilient infrastructure 100% renewable electricity PUE 1.2 in Munich AI Factory Net zero emissions since 2025 (Scope 1 & 2) 2024 – Green AI Principles for coding and infrastructure Heat re-use & intelligent cooling in Data Centers Hardware circularity – ensuring availability and management of IT and network resources Crisis management – being where we are needed Built-in network resilience – redundancy, power and recovery Intelligent protection of network and services against extreme weather and physical threats Business continuity by design AI-powered Cybersecurity at scale AI-analysis of >5bn security-relevant datapoints per day AI-powered detection and defense of >20k DDoS attacks per month 99% accuracy in AI-detection of false positives (first level support) European sovereignty – trusted data and AI operations Secure-by-design – building trust from development to operation Scale AI with efficient energy, compute and resources Enable safe and reliable AI use for customers and business Ensure continuity when systems, sites and supply are under stress
Page 8
T oday: AI is not a vision anymore 8 AI Innovation Hub Joint lab researching future AI use cases e.g. AI RAN >95% faster incident handling RAN Guardian Building the network of the future: 6G Joint initiative to design and standardize the future AI-native network T-Challenge Joint events to discover and interact with the AI ecosystem ~42% Code AI generated Code 2.6 mn deflected calls / H1 ‘26 AI driven Bot >83% AI usage Employees use AI ~10k NVIDIA GPUs Industrial AI Cloud -55% # Calls Call reduction +12% NPS Expert Assist ~32% Contacts handled AI Agent >30 k adjustments Antenna adjustments during Winter Storm Fern 400+ B2B Use Cases, e.g. Voice AI Notes, Agentic Hub, “Schockraum” Use Cases
Page 9
100% calls initially handled by AI 50–55% reduced calls & chats (vs 2023) Up to 70% retention & tariff changes via digital +5 tNPS Sales & Service touchpoints AI Transformation operationally steered via 2030 AI ambitions DT ex US; selected 2030 examples 9 Customer interface Network It & Software Financial Impact 10–15% Network gross savings1 Level 4 Autonomous Network for high value use cases >4 CXI for >90% of mobile customers + 50–60% software developer productivity 70–80% of our code written with AI 15–20% IT gross savings1 ~€1.1 bn by 2027 Gross savings from AI and automation2 ~€2.5 bn by 2030 Gross savings from AI and automation1 >€ 1 bn Additional revenues from new AI value pools by 2030 1 IDC AL, 2030 vs. 2023, DT ex US, excl. token costs, incl. cap. labour 2 vs 2023, IDC AL and Capex, excl. token costs
Page 10
AI reinforces our successful flywheel strategy 10 Financials Enhancing consistent growth, higher margins and new revenue pools Efficiencies AI-powered operations across every single workflow Our T: People & Brand Growing our people for AI-native execution & turning our trusted brand into an AI advantage Investments A network powering Physical AI with low latency services (6G) & sovereign AI Clouds Customers Hyperpersonalized, intelligent, instant experiences and products Global scale Leveraging transatlantic scale in the AI economy Data & AI Accelerate every turn