The latest survey on the state of AI adoption among Vietnamese businesses shows that generative AI (GenAI) has become widely used at the individual level. However, only 15% of respondents reported increased revenue, highlighting a significant gap between AI adoption and its actual business value.
According to a survey conducted by 3HORIZONS in collaboration with the CIO Vietnam community, around 95% of respondents said GenAI is currently being used at the individual level.
However, this figure drops significantly when looking at deeper levels of integration. Only 59% have incorporated AI into their workflows, while 36% have embedded AI into business applications or processes. For more autonomous forms of AI, the figure falls to just around 7%, although 44% of businesses have already experimented with or are using AI agents.
The survey suggests that the key issue lies not in any individual figure, but in the gap between AI usage and the ability to scale it. Employees using AI more frequently does not necessarily mean that the technology has been integrated into core business processes or is delivering tangible business results.
Dr. Tran Viet Huan, Chairman of CIO Vietnam, said: “AI is already being used very widely, but widespread use does not mean enterprise-wide deployment, and scale does not necessarily translate into business value.”
According to Mr. Huan, a key question now is how to turn AI from a collection of individual tools and experiments into a scalable, measurable, and accountable organizational capability.
The gap becomes even more apparent when looking at actual business outcomes. Around 90% of respondents reported that AI had improved employee productivity, while 62% said it had supported better decision-making and 52% reported an impact on innovation. However, the figures drop sharply for metrics directly linked to financial performance: 33% reported cost savings, 23% saw improvements in customer experience, and only 15% reported revenue growth.
In other words, AI is currently delivering much stronger improvements in individual productivity than measurable impacts on costs, revenue, or business models — and this is the challenge now facing business leaders.
According to the survey, the main barriers to putting AI into practical operation are organizational capabilities, including access to reliable data, the ability to integrate systems, scalable infrastructure, AI-skilled talent, coordination between business and technology teams, change management, security and privacy, as well as the ability to control costs and measure effectiveness.
Google translate