Southeast Asian tech leaders are learning to trust AI agents, but not with production
AI coding agents have moved quickly from novelty to daily tool inside Southeast Asian engineering teams. Across Singapore, Bangkok, Jakarta and Ho Chi Minh City, developers are no longer just asking AI to complete a line of code or explain an error message. They are using agents to plan tasks, refactor repositories, generate tests and […] The post Southeast Asian tech leaders are learning to…
Southeast Asian engineers are embracing AI coding agents as daily tools, using them to plan tasks, refactor code, generate tests, and execute multi-step workflows. However, as these tools become more capable, a crucial question arises: how much autonomy is too much? The Agoda AI Developer Report 2026 surveyed over 800 developers and engineering leaders across Southeast Asia and India, revealing that 53% of developers have deployed AI agents in production or organizational use.
While this adoption is significant, developers remain cautious when AI touches live environments, with 79% requiring human approval before AI-generated work can be pushed live. The cautious approach stems from the fact that AI systems still produce hallucinations, incorrect outputs, and overconfident answers, which can be costly to catch and reverse.
Therefore, developers put human approval gates around decisions that could break systems, expose security gaps, or affect customers.
Written by urgent.news from e27's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
