Southeast Asia's developers self-teach AI faster than companies can train them

A new report from Agoda reveals a workforce racing ahead of itself. Across Southeast Asia and India, developers are teaching themselves artificial intelligence faster than their employers can organize formal training programs—a dynamic that is reshaping how technical talent grows, but also exposing deep fractures in who gets access to what.

The numbers tell the story plainly. Eighty-seven percent of developers in the region have already shifted their learning priorities or career direction because of AI. Sixty-two percent believe the technology will create more jobs than it eliminates. Yet when it comes to how they're actually acquiring these skills, the picture is lopsided: seventy-two percent are self-taught, while only twenty-eight percent receive structured training from their employers. The gap is not a minor inefficiency—it suggests that the region's tech workforce is advancing on its own terms, driven by individual initiative rather than institutional planning.

Peer learning has become the backbone of this self-directed education. More than half of developers—fifty-two percent—rely primarily on online communities and open-source projects to learn. These are not formal classrooms or corporate curricula. They are spaces where engineers collaborate, experiment, share code, and teach each other in real time. The culture is fast-moving and collaborative, but it is also unstructured and dependent on whoever happens to be present and willing to help.

This momentum, however, is not evenly distributed across the region. Developers in Singapore are nearly twice as likely as their counterparts in Vietnam to have access to formal AI training. The gap between senior and junior engineers is even starker—a twenty-five-point confidence gap separates the two groups, suggesting that mentorship and structured support can meaningfully accelerate skill development. Those with experience can move faster; those without it are left to catch up on their own.

The pressure is mounting. Forty-four percent of developers worry about falling behind as AI technology evolves. Fifty-eight percent now view AI proficiency as a baseline requirement for hiring, not a nice-to-have skill. The result is a workforce that is ambitious and adaptable, learning faster than organizations can respond, but also strained and uneven. Junior engineers in Vietnam face different odds than senior engineers in Singapore, even though both are racing against the same technological clock.

Agoda's Chief Technology Officer, Idan Zalzberg, framed the challenge directly: developers across the region see AI as a tool to accelerate their work, not as a replacement for human judgment, and they are upskilling themselves faster than many organizations can respond. The self-driven momentum is powerful, he said, but sustaining it requires companies to build systems of trust and accountability around their engineers. Agoda itself has invested in several initiatives—internal AI hackathons where engineers experiment with emerging technologies, regular tech talks and knowledge-sharing sessions to accelerate peer learning, and Tech Camp Day, a program that has reached thousands of university and high school students across Thailand with foundational technical skills.

The report draws on input from developers across Southeast Asia and India, along with insights from regional companies including Carousell, MoMo, Omise, and SCB 10x. It offers a snapshot of how AI is being adopted and experienced across the region's developer ecosystem at a moment when the technology is reshaping what it means to be a software engineer. The question now is whether companies and institutions can build structures fast enough to support the learning that is already happening on its own.