Agoda’s AI Developer Report 2025 shows that AI has become a mainstream tool for developers in Southeast Asia and India, delivering real productivity gains while raising new questions about reliability, skills, and organisational readiness.
The report finds that AI adoption is nearly universal: 95% of developers in Southeast Asia and India use AI tools weekly, and more than half keep an AI assistant open while they work. Developers report clear speed improvements—over 90% say AI makes them faster—but describe the impact as steady, incremental gains rather than dramatic automation of their jobs.
Despite high usage, developers are pragmatic in their reliance on AI. Productivity is the primary driver: 80% cite speed and help with routine coding tasks as key benefits, yet most still treat AI as an assistant to be supervised rather than an autonomous actor. Agoda summarises this as “AI is mainstream, but not mature,” with workflows and safeguards still catching up to the tools.
One of the clearest gaps in the report is between developer behaviour and organisational governance. Tooling has consolidated—ChatGPT is used by around 87% of respondents, and many say their IDEs are already “AI-ready”—but formal policies lag. Only about one in four teams report operating under official AI guidelines, and roughly 60% say their organisation has no formal AI policy in place.
In practice, developers are filling this gap with bottom-up practices. Reliability and inconsistent outputs are named as the top barrier to broader use by nearly 80% of respondents, yet 67% say they constantly review AI-generated code before merging, and around 70% routinely rework AI output to ensure correctness. The report argues that this “review-by-default” culture is turning high adoption into responsible adoption, even before top-down frameworks are fully established.
The study highlights a strong push toward self-directed learning. Around 87% of developers say they are self-training on AI, using online materials, experimentation, and peer learning rather than formal company programs. At the same time, expectations are rising: more than half believe AI proficiency should be a hiring requirement, and many see AI skills as essential for career progression.
Agoda notes that developers across the region view AI as a way to augment rather than replace their work, but that this bottom-up momentum can expose gaps in access and support. Companies are now challenged to catch up with their own developers by offering structured training, clearer policies, and frameworks that align experimentation with security, compliance, and long-term capability building.
The report describes three emerging realities: AI is mainstream but unevenly distributed, evolving through accountability, and experience with AI varies widely across markets and companies. Most developers report time savings in the range of one to six hours per week, with the most significant impacts on coding, debugging, and learning new APIs or frameworks, rather than full task automation.
Agoda’s CTO, Idan Zalzberg, frames this as a shift in how software is built rather than who builds it: AI is becoming embedded in everyday workflows, from code review to collaboration, but human judgment and oversight remain central. The report concludes that the real opportunity for the region lies in pairing this ground-up experimentation with stronger systems of trust, accountability, and shared practices—turning near-universal adoption into a durable, region-wide capability.
