Organizations in Southeast Asia are adopting observability tools for artificial intelligence faster than global peers, but engineering teams in the region are still losing a large share of their working capacity to system disruptions, according to New Relic’s 2026 Observability Forecast.
The ASEAN findings cover 150 IT and engineering leaders and practitioners in Singapore, Indonesia, Malaysia and Thailand. New Relic said 66 percent of respondents in that group have deployed AI application observability, compared with 47 percent globally, while 59 percent use AI-assisted troubleshooting.
The regional data form part of a broader survey conducted with Enterprise Technology Research covering 2,575 respondents across 24 countries and 12 industries. New Relic’s Asia-Pacific edition includes 950 respondents across India, Japan, Australia and New Zealand, and Southeast Asia.
Agentic AI becomes a major observability spending driver
New Relic said 39 percent of ASEAN respondents cited adoption of agentic AI systems as a driver of observability spending, the highest rate among the groups included in its comparison. The company also said 86 percent of ASEAN respondents agreed that growth in AI-generated code makes observability more important.
The figures come as enterprises move from monitoring conventional applications to tracking autonomous or semi-autonomous software that can make changes, invoke tools and interact with production systems. Globally, New Relic found that one quarter of organizations have deployed AI agents in production without monitoring them through observability tools.
For Southeast Asian companies, the operating challenge is not only whether AI workloads are visible, but whether signals from applications, infrastructure and agents can be correlated quickly enough to reduce incident-response time.
Engineering teams still spend heavily on firefighting
Despite relatively high adoption of AI monitoring, ASEAN engineering teams said they lose a median 47 percent of their capacity to system disruptions, compared with 37 percent globally. New Relic also reported that 10 percent of ASEAN organizations experience high-impact outages multiple times a day, up from 1 percent in the previous year’s survey.
The survey put the average cost of a high-impact outage among ASEAN respondents at $1.62 million, with a mean time to resolve of 52 minutes. Those figures are survey-based estimates and should not be treated as audited financial-loss data across the entire region.
New Relic also identified a gap between use of AI-assisted troubleshooting and more basic operational practices. It said 23 percent of ASEAN respondents still detect critical outages primarily through manual checks performed at intervals, while 47 percent cited technology-stack complexity as a barrier to full-stack observability.
OpenTelemetry adoption rises while tool fragmentation persists
Thirty-three percent of ASEAN respondents said their organizations are fully standardized on OpenTelemetry, compared with 22 percent globally, while another 27 percent are actively migrating. OpenTelemetry provides a common framework for collecting and exporting telemetry such as traces, metrics and logs.
Standardizing telemetry does not by itself eliminate fragmentation. New Relic said 53 percent of ASEAN respondents still prefer using multiple point tools, which can leave teams correlating alerts and operational data across separate systems during incidents.