Only 2% of 55 Asia-Pacific telco respondents reported live agentic AI at scale
TM Forum's wider survey covered 216 IT executives from 111 operators in 72 countries. Its 55-person Asia-Pacific subset points to a gap between strategy and production use—but the self-reported regional result is directional, not an audited market estimate.
The short version
- What happened
- TM Forum says 2% of 55 Asia-Pacific survey respondents reported that AI agents were operating at scale in live core systems; 72% described AI as experimental or attached to specific tasks.
- Who it affects
- Telecom workers may gain tools that diagnose faults or assemble customer information, but live agents can also change roles, decision authority and accountability. Customers will feel the result through service reliability, complaint handling, privacy and whether a human can intervene when automation is wrong.
- What remains unproven
- The Asia-Pacific analysis has 55 respondents and is explicitly described by TM Forum as directional, not statistically definitive.
Keep in mind
The Asia-Pacific analysis has 55 respondents and is explicitly described by TM Forum as directional, not statistically definitive.
The public release does not disclose fieldwork dates, response rate, sampling frame, weighting or the distribution of respondents by operator and country.
Responses are self-reported and mix executive, architecture and engineering perspectives; the study does not independently audit live agents or measure incidents, productivity, service quality or customer outcomes.
The reported 2% is rounded, so it should not be converted into an exact operator count without the underlying data.
Source analysis · 1 linked source. Primary analysed source date: 28 October 2026.
Keep following the evidence
Join The Impact Brief founding list for selected AI reports, their effect on people and the evidence limits. Read a sample first.
Read the full report · 4 min The evidence, methods, implications and limitations
The production gap is larger than the strategy gap
Telecommunications companies increasingly describe AI as central to network operations and customer service, yet TM Forum's Asia-Pacific results suggest that production-scale agents remain unusual. Among 55 regional respondents, 89% called AI critical to competitive strategy, while 2% reported AI agents operating at scale in live core systems. A much larger 72% said their use remained experimental or bolted onto particular tasks, and 55% placed core-enterprise integration among their top three priorities for the next 18 months.
Those figures do not show that agent projects have failed. They show different stages of adoption being discussed in the same market. A chatbot, a technician copilot and an agent allowed to change a live network carry very different operational consequences. Moving from recommendation to action requires reliable data, documented decision rules, identity and access controls, observability, rollback and a clear point at which a person must intervene.[1]
The survey identifies barriers, not their measured cost
Regional respondents most often cited shortages of skills as a major obstacle, at 49%, followed by low-quality data at 43%, cost at 42% and accuracy or hallucination concerns at 39%. Another 38% said critical human decision-making was not documented well enough to scale AI safely and consistently. That last finding is practical: an agent cannot reliably reproduce or escalate a judgement that the organisation itself has never made explicit.
The denominator also limits the claim. The global research surveyed 216 IT executives from 111 operators across 72 countries, including C-suite leaders, network architects and software-engineering practitioners. The public release does not give fieldwork dates, a response rate, sampling frame, operator-size distribution or weighting, and respondents may not represent every operator. TM Forum itself describes the 55-person Asia-Pacific view as directional. Stronger evidence would independently verify deployed agents, the tasks they perform, intervention and rollback rates, incidents, customer effects and productivity over time.[1]
The distinction that matters
What 55 Asia-Pacific respondents reported
The percentages describe organisational views and claimed adoption stages, not audited market shares.
2%
Reported live agentic AI at scale in core systems
Rare in this directional regional sample
72%
Described AI as experimental or limited to specific tasks
Most reported activity remained before broad core integration
49% / 43% / 42%
Cited skills, data quality and cost as major obstacles
Organisational foundations may constrain deployment more than model access
Watch the background, then check the evidence
Illustrative photo · not a video frame
UNESCO · 11 min 27 sec · Background explainer
UNESCO Talks: Interview with 2024 Nobel laureate Professor Daron Acemoglu
YouTube loads only when you choose; playback starts when you press play.
Economist Daron Acemoglu discusses how choices about AI affect workers, inequality and society. Pair the interview with our comparison of task exposure, hiring and skills evidence.
Evidence check: An expert interview from 2024, offering a perspective on policy and work rather than a measured forecast of job losses.
Published 2024-10-14. Source checked 14 October 2026. Check captions on the publisher’s video page.
Read the AI and jobs guidePhoto: Dylan Gillis / Unsplash · Licence
Continue the story
Read next
Continue exploring enterprise adoption: compare the finding and its evidence limits.
Do agentic tools now make up most business AI use in the EU?
Within OpenAI's Enterprise and Business customer sample, yes: Codex and ChatGPT Work produced 55% of included output tokens at the end of September. That is not a measure of all EU firms, completed work or productivity.
Read the next reportExplore the full source and evidence record
What this means for people
- Telecom workers may gain tools that diagnose faults or assemble customer information, but live agents can also change roles, decision authority and accountability. Customers will feel the result through service reliability, complaint handling, privacy and whether a human can intervene when automation is wrong.
Global context
Telecom operators are attractive test beds for agentic AI because their networks produce extensive operational data and require rapid fault response. They are also high-consequence environments: mistakes can affect emergency access, payments and public services. The survey supports cautious investment in data, skills and controls before autonomy—not a race to maximise the number of agents in production.
What the evidence does not yet show
- The Asia-Pacific analysis has 55 respondents and is explicitly described by TM Forum as directional, not statistically definitive.
- The public release does not disclose fieldwork dates, response rate, sampling frame, weighting or the distribution of respondents by operator and country.
- Responses are self-reported and mix executive, architecture and engineering perspectives; the study does not independently audit live agents or measure incidents, productivity, service quality or customer outcomes.
- The reported 2% is rounded, so it should not be converted into an exact operator count without the underlying data.
What to watch next
- The complete survey methodology
- Country and operator-level breakdowns
- Independent verification of live deployments
- Human-intervention and rollback rates
- Customer and network reliability outcomes
- Measured costs and productivity after deployment
Living evidence record
Impact record IAI-0Z43IUK
Evidence stage
Announced
Confidence
Developing
Reporting basis
Source analysis
Independent or research support
Not yet
Record status
Monitoring
Last checked
31 October 2026
Source trail
1 direct source across 1 source type.
People impact
Documented in this record.
Uncertainty
Limits and next checks are explicit.
Stages describe the evidence available—not whether a technology is good or bad. See the public method.
Single-source reporting disclosure
This record analyses one direct source. It can establish what TM Forum published or reported, but it is not independent corroboration of every performance claim or predicted outcome. The confidence label will change only when broader evidence is added.
Evidence trail
Sources used for this report
Links checked 31 October 2026
Research question, topics and regions
Topics: Telecommunications · Agentic AI · Enterprise adoption · Skills · Data quality · Operational governance
Regions: Asia · Oceania · Asia-Pacific · Global
This report is labelled source analysis. We summarise and analyse source material in our own words; company statements remain attributed claims until independently supported. Translated summaries preserve the meaning of the original source and link back to it. Read our editorial standards.
Reader commentary
What does this mean where you are?
Tell us what you have seen at work, in your community or in everyday life. A question is welcome too. No account is needed; we review notes before publishing them.
Explore commentary across the portal →Published reader notes
0No published reader notes yet. You can start the evidence-led discussion above.
Prefer a private correction or response? Contact the newsroom.