Adoption is not capability – the real AI test for Singapore’s SMEs
The Republic’s AI scoreboard looks strong. But, in many firms, the whole capability rests on one person
Singapore is positioning itself as a regional leader in artificial intelligence (AI) adoption, but its success hinges on more than just individual champions. A recent study commissioned by Amazon Web Services (AWS) found that while 75% of financial services SMEs, along with a majority in healthcare and manufacturing, are using AI, only 60% of firms have deep, sustained capability.
In many cases, the removal of a single person leads to AI work stalling or stopping entirely. The real test for Singapore's SMEs is whether this capability can endure beyond any one individual. At the heart of this challenge is the translation of AI outputs into meaningful business value. Often, the expertise lies not with the AI engineers but with domain experts who understand the intricacies of their industries.
When these experts are removed from the equation, the AI's potential is lost. AWS has recognized this by committing $1 billion to deploy forward-deployed engineers within customer teams, empowering them to self-sufficiently manage AI solutions. However, the shift from individual expertise to team fluency is crucial for sustainable adoption.
While training programs and showcases of successful implementations are valuable, they do not fully bridge the gap. The true difference-maker lies in practical, hands-on experiences where SMEs encounter and learn from the challenges of integrating AI into their operations. Singapore's National Business Survey 2025 reveals that about a third of firms still harbor doubts about AI's accuracy and reliability, highlighting the need for a deeper understanding.
To foster lasting capability, Singapore should focus on creating platforms where SMEs can share their actual learnings, both successes and failures, without the fear of judgment. By nurturing fluency through user-level engagement and providing domain experts with the necessary tools and sandbox environments, Defa can become a robust mechanism to transfer proven AI capability across the region.
Written by urgent.news from The Business Times - Singapore's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
