- Telecom sector leaders gathered at the TM Forum’s Innovate Asia event in Bangkok
- The industry body’s leadership noted that Asian telcos are setting the pace as operators migrate towards autonomous networks and AI-native operations
- But with opportunities come challenges, especially as AI agents interact with telco systems
BANGKOK – Innovate Asia 2026 – Asia is emerging as a driving force in the transformation of the global telecom sector, with operators in the region making significant progress towards autonomous networks (AN) and AI-native businesses.
That was the central message from TM Forum’s leadership here in Bangkok, where industry executives gathered to examine the progress being made in moving from basic network automation to greater autonomy, and the opportunities this transition presents for operators.
Opening the event, TM Forum chairman Steffen Roehn highlighted Asia’s pivotal role in defining the future direction of telecom. “Eleven of the top-20 operators are headquartered in this region,” he noted, pointing to the scale, diversity, investment and pace of innovation that make Asia a natural centre for the industry’s transformation.
For Roehn, the region is not simply adopting developments taking place elsewhere – it is actively shaping them. “When people want to know what the future of telecom looks like, when they want to see AI-native networks running in production, autonomous operations delivering real business value, and architectural decisions being made at scale, they look here.”
He also emphasised the importance of ensuring that innovation developed in Asia benefits the wider global industry, and pointed to TM Forum’s platform through which frameworks, code, implementation experience and practical solutions can be shared internationally.
The conference theme, ‘From Architecture to Autonomy: Asia’s Race to 2030’, reflects the industry’s fundamental shift from experimentation to implementation, with a growing emphasis on translating AI and autonomous capabilities into measurable business results.
Autonomous networks move from ambition to reality
The scale of the progress being made across the region was reinforced by Nik Willetts, president and CEO of TM Forum, who pointed to a significant increase in operator commitments and achievements over the past year.
Almost 40% of the world’s service providers now have ambitious targets for autonomous networks, compared with just 25% a year earlier, according to the forum’s research. More importantly, the number of Level 4 autonomous network certifications has jumped from just 15 across six service providers to 53 across 20 service providers. Significantly, 87% of these achievements have come from Asia. “When I say you’re setting the pace, it’s not just to be kind. This is real, and this is having real results,” said Willetts.
That momentum is borne out by the latest list of telcos to have achieved autonomous networking validations, with Thailand’s True, South Korea’s KT Corp. and Vietnam’s Viettel the latest to have achieved AN Level 4.
The significance of this progress extends beyond operational efficiency. While early autonomous network initiatives were largely driven by the need to reduce costs and improve network performance, operators are now demonstrating tangible benefits for customers, shareholders and employees alike.
These include lower energy consumption, improved customer experience and the ability to identify and resolve service issues autonomously. However, Willetts believes the industry is approaching a much bigger opportunity: Redefining what telcos sell and how they differentiate themselves in an AI-driven economy.
From connectivity to customer outcomes
Traditional connectivity services are increasingly commoditised, placing pressure on operators to find new sources of value rather than competing solely on price. The opportunity, according to Willetts, lies in recognising that customers are less concerned about the underlying technology than the outcomes it enables. “Consumers want to make sure they’re always connected, because their entire life now works online,” he noted.
For enterprises, the requirements are even more demanding. As businesses begin to rely heavily on AI to manage their operations, they will need connectivity and network services that can deliver guaranteed, measurable and trustworthy outcomes. This creates opportunities for operators to move beyond conventional connectivity packages towards more tailored services, with pricing models based on performance, reliability and specific business requirements.
Willetts highlighted the enduring importance of service-level agreements (SLAs) in this emerging environment. TM Forum research indicates that 70% of enterprises identify SLA compliance as a priority, with network reliability and performance ranking among their top requirements. Yet only around one-third are satisfied with the services they currently receive. This represents a significant commercial opportunity for operators willing to make stronger commitments to their customers.
The ability to dynamically scale network performance according to demand, whether in a densely populated urban environment or a remote location, will become vital. Equally, the boundaries between terrestrial and satellite connectivity will need to become less visible to customers. “Our customers don’t care how they’re connected. They just want to be connected all the time,” observed Willetts.
The development of integrated connectivity, combined with the intelligence provided by autonomous networks, could allow operators to deliver seamless services that respond dynamically to customer requirements.
Networks become part of the intelligence infrastructure
Beyond connectivity, another significant opportunity lies in the evolution of networks into intelligent infrastructure. Network sensing capabilities – already emerging through 5G and expected to develop further as 5G Advanced and 6G networks are deployed – could allow operators to gather information not only about the digital environment but also about the physical world. This will prove crucial as physical AI expands, creating new requirements for systems that can interact intelligently with their surroundings.
Rather than allowing AI infrastructure to develop independently of telecom networks, Willetts believes operators should position themselves as an integral part of the intelligence infrastructure supporting the AI economy. The industry’s existing assets, including its networks, local market knowledge and trusted relationships with governments, businesses and consumers, provide a potentially important competitive advantage.
He also sees an increasingly important role for telcos in protecting customers and national infrastructure. As sovereign entities operating critical national infrastructure, operators are uniquely positioned to support trusted and secure AI services, provided they make security and trust central to their future business models.
The challenge of the agentic customer
Perhaps one of the most significant changes facing telcos is the emergence of AI agents as a new channel through which customers will interact with businesses.
Consumers are already using AI assistants to manage aspects of their personal lives, including finances, bills and subscriptions. In this new landscape, telecom services will increasingly be managed by AI agents acting on behalf of customers. “In their world, we are a bill and a subscription,” Willetts observed.
This presents a fundamental challenge for an industry whose customer service operations, commercial systems and business processes have traditionally been designed around human interaction.
Willetts believes operators need to prepare for a future in which customers expect to interact with telcos through AI agents, potentially within the next 18 months. The implications extend well beyond adding an agentic interface to existing systems. Operators will need to establish how agents are authenticated, how their authority to act on behalf of customers is verified, and how transactions and decisions can be securely managed.
Likewise, enterprise customers will expect their own AI agents to provision services, adjust capacity and change requirements dynamically in response to business needs. For operators, failing to embrace this transition risks creating opportunities for new intermediaries to take control of the customer relationship.
“If we ignore that opportunity and just see it as a threat to be put to one side, the reality will be that somebody else will disintermediate us,” warned Willetts.
The challenge is, therefore, not simply a technological one. It requires a fundamental rethink of how telco businesses are structured and how their services are delivered.
Building the foundations for autonomous enterprises
TM Forum has been working with its global membership to address these challenges, moving beyond individual AI use cases towards a more comprehensive architectural approach.
Willetts argued that implementing isolated AI agents within existing business structures risks creating new silos, with separate systems making independent decisions that cannot consistently deliver end-to-end customer outcomes.
In response, TM Forum has introduced the concept of ‘autonomous flows’ – closed-loop processes designed to connect customer intent with trusted outcomes. The objective is to create systems capable of sensing, deciding, acting, learning and governing, with the ability to deliver the required outcome reliably and consistently. Achieving this rests on three fundamental principles.
First, business capabilities must become composable, allowing operators to combine and reconfigure services and processes to create new value streams.
Second, autonomy must extend across the entire business process rather than remain confined to individual functions or network domains. Systems must also continuously learn and improve.
Third, trust must be embedded throughout the architecture, ensuring that AI-driven decisions are transparent, accountable and governed appropriately.
The foundations supporting this transformation include trusted, contextualised data, governed data products, knowledge graphs and digital twins. They also require AI-discoverable business capabilities, building on TM Forum’s Open Digital Architecture (ODA) components, alongside open application programming interfaces (APIs) designed to support AI-driven interactions. Shared semantics and a common understanding of customer and business context will be essential if agents are to communicate effectively across different systems and organisations.
Equally important will be secure agent-to-agent communication, appropriate AI standards, and the ability to monitor, govern and audit agent behaviour at runtime. TM Forum’s global community, comprising more than 800 companies and 100,000 engineers, is working to establish these foundations and accelerate their adoption across the industry.
Trust becomes a competitive differentiator
For telcos, trust could become one of their most valuable competitive advantages in the AI economy. Willetts believes operators occupy a unique position thanks to their established presence in local markets, their relationships with governments and their responsibility for critical national infrastructure. “Trust is more than an obligation. For me, this industry should treat trust as a licence to operate,” he stated.
However, maintaining that position will require operators to understand how AI models and agents behave, the infrastructure on which they run, and the implications of changes to underlying models.
TM Forum is exploring frameworks such as model-as-a-service to provide greater visibility into the origins, configuration and operational characteristics of AI models. This includes understanding what happens when an underlying model changes and how that change might affect the behaviour of agents and the decisions they make.
For a regulated industry responsible for critical infrastructure, the ability to establish accountability and confidence in autonomous decision-making will be fundamental.
Collaboration will determine the winners
While the progress made by Asian operators is impressive, the next stage of transformation will require greater collaboration across the global telecommunications ecosystem. Roehn challenged delegates to build on the achievements already made, increasing the number of engineers capable of implementing advanced autonomous network use cases. Recalling the previous year’s event, when engineers received certification for Level 4 autonomous network implementations, he expressed his ambition to see the number grow from 40 to 400 and, ultimately, 4,000.
For Willetts, the industry’s future will depend on its ability to anticipate competitors and business models that may not yet exist. “The race won’t be won by each company trying to compete with what we see today as our competitors. It will be by understanding that our competitors may not even exist yet,” he stressed.
The first autonomous flows were being demonstrated live at Innovate Asia 2026, providing a practical illustration of how the concepts discussed by industry leaders are beginning to move from architecture and experimentation into implementation.
The message from Bangkok was clear: The transition towards the autonomous, AI-native telco is no longer a distant ambition. The technology is maturing, the business case is becoming clearer, and operators in Asia are already demonstrating measurable progress. The challenge now is to scale that progress, develop new sources of value, establish trusted AI foundations and fundamentally rethink how telcos operate.
As Willetts concluded, the industry will need to collaborate to build those foundations. “This race will be won together,” he said.
- Tony Poulos, Contributing Editor, TelecomTV
Email Newsletters
Sign up to receive TelecomTV's top news and videos, plus exclusive subscriber-only content direct to your inbox.