Vietnamese businesses report AI gains, but most are still experimenting

Vietnamese firms report rising productivity and revenue gains from AI, while most adopters are still experimenting rather than scaling it across their operations. A study presented by AWS puts numbers on both the early returns and the distance still to travel.

AWS AI Watch analysis

What happened

The study, conducted by Strand Partners and based on surveys of 1,000 business leaders and 1,000 members of the public, found that 26 per cent of Vietnamese businesses had adopted AI. Among those adopters, 72 per cent reported productivity gains and 64 per cent said AI had boosted revenue, with an average increase of 15 per cent. These are survey findings, not audited company accounts.

The same study found that 61 per cent of AI adopters were still experimenting. Of businesses familiar with agentic AI, 8 per cent said they had embedded AI agents in core workflows, while 19 per cent were piloting or experimenting with them. AWS also highlighted individual deployments: securities firm TCBS said more than 460 employees use its Kiro coding environment, while VPBank is developing an AI platform with AWS and OneByZero. Read Vietnam Investment Review’s report.

Why it matters

The gap between reported benefits and widespread deployment is the useful signal. AI may be delivering value for some Vietnamese organisations, but the numbers do not describe a market that has finished working out where the technology fits. The study says 78 per cent of adopters expect AI-driven growth in the next year; that is an expectation, not a result already in the bank.

The examples also show what “adoption” can mean in practice, from software development to a shared platform for building AI assistants. They are concrete cases, though the report does not provide independent performance comparisons for either deployment.

Our read

This is a worthwhile snapshot of AI moving from trial projects towards everyday business use, with a reminder that enthusiasm and measurable returns are not the same thing. The numbers come from a study presented by AWS, and should be read as survey responses rather than a neutral census of the whole economy. Still, the split between experimentation and embedded use is more revealing than another forecast about AI changing everything by Tuesday.

What to watch

• Whether the share of businesses using AI, and the share reporting gains, rises in later surveys.
• Whether pilots of agentic AI become embedded in core workflows.
• Whether the named business deployments publish measurable results, not just adoption figures.

Discussion spark: Should businesses scale AI once staff report productivity gains, or wait for independently measured returns before making it part of core operations?

not affiliated with or endorsed by Amazon Web Services (AWS)