TM Forum members unveil AI-native ODA roadmap for autonomous telecoms

TM Forum members are introducing AI-native extensions to Open Digital Architecture ahead of DTW Ignite 2026 in Copenhagen, pairing a roadmap, live demos and a new report to help telecom operators move from AI pilots to trusted autonomous operations. The effort is meant to give communications service providers a shared foundation for scaling AI across IT and network operations. Why it matters: - The telecom industry is trying to move from scattered AI experiments to autonomous operations that can scale across business and network systems. - TM Forum’s AI-native ODA effort is meant to give communications service providers a common architectural foundation for that shift. - The roadmap is tied to TM Forum’s Race to 2030 strategy, which links autonomy to stronger performance, reliability, sovereignty and assured outcomes. What happened: - TM Forum members are preparing to introduce AI-native extensions to the Open Digital Architecture at DTW Ignite 2026 in Copenhagen, set for June 23-25. - The launch package includes an industry-backed roadmap, live Level 4+ demonstrations and a new TM Forum Insights report. - TM Forum says the materials are designed to help communications service providers move from fragmented AI pilots to autonomous flows at scale. The details: - The AI-native extensions add a governed execution layer for AI agents inside ODA. - The architecture builds on TM Forum’s component architecture and Open APIs, with the ODA Canvas providing control points. - The approach keeps AI working alongside trusted systems of record, not in place of them. - TM Forum says that design preserves control, accountability, interoperability and architectural integrity as autonomy grows. - New AI-Native ODA capabilities include ODA Canvas extensions for automated lifecycle management of AI agents and integration with ODA Components via MCP. - The new capabilities also include ODA Canvas Operators for centralized management and control over AI, secure agent interactions, guardrails and managed access to data products and large language models. - A reference implementation supports demonstrations of Level 4+ autonomy on the AI Native ODA Canvas foundation. - TM Forum’s roadmap is set to be endorsed by members in Copenhagen. - The roadmap outlines a path to trusted, AI-native operations at scale and supports the broader Race to 2030 strategy. - The roadmap focuses on moving from isolated AI use cases to autonomous flows that connect intent, orchestration, assurance and action across business and network operations. - TM Forum’s industry Missions on Composable IT and Ecosystems, Autonomous Networks, and Trustworthy AI and Data will support implementation across IT and networks. - TM Forum Insights has published “Transformation 2.0: Towards an AI-native Open Digital Architecture.” - The report says ODA is evolving from a cloud-native foundation into the architectural backbone of the AI-native telco and autonomous enterprise. - The report identifies ten interconnected challenges that CSPs need to solve to deploy generative AI and agentic AI at scale. Between the lines: - The push signals that AI in telecom is moving from isolated use cases toward governed operational frameworks. - TM Forum is positioning standards, reference code and shared operating rules as the missing layer between AI experimentation and production autonomy. - The emphasis on sovereignty, guardrails and deterministic systems suggests the industry wants AI adoption without sacrificing control in regulated network environments. What’s next: - TM Forum members are expected to endorse the AI-native ODA roadmap in Copenhagen. - The Industry Showcase at DTW Ignite will demonstrate end-to-end fault management, dynamic 5G network slicing and sovereign AI inference. - The demonstrations are intended to show Level 4+ cross-domain autonomous flows in real operational scenarios. - TM Forum says the roadmap, capabilities, demonstrations and report will work together to guide how CSPs build, govern and scale AI-native autonomy.