Despite lofty goals to transform businesses, about 20 per cent of AI projects in Asia-Pacific end up failing because of poor data quality, inherent bias and complex engineering processes, according to research firm IDC.
In a report released this week, it also revealed that 80 percent of Asia-Pacific organisations it recently surveyed were re-evaluating their data management strategies.
The study, commissioned by Qlik and and co-sponsored by Amazon Web Services (AWS), found that inconsistent, outdated, and biased data not only undermines trust, but also hinders effective decision-making.
Some 35 per cent of organisations surveyed cited these challenges as a key barrier to successfully scaling their generative artificial intelligence (GenAI) initiatives.
These insights align with findings from recent research conducted by Hitachi Vantara and other business technology firms, pointing to the need for strong data foundations to support AI projects.
Despite these challenges, AI-related investments in the region are expected to exceed overall digital technology spending by 1.7 times, with GenAI expected to have an estimated regional economic impact of US$1.6 trillion by 2027, according to IDC.
The majority of organisations across the Asia-Pacific region, including Japan, are still in the early to mid-stages of adopting AI and machine learning (ML), it found. Only 24 per cent have managed to successfully scale and embed these technologies into their core business operations.
To address these challenges, 56 per cent of the Asia-Pacific organisations surveyed are turning to low-code and no-code AI tools to streamline the implementation of GenAI.
Two key priorities for AI are having a strong data foundation and identifying the right business use cases, said Dr Chris Marshall, IDC’s vice-president for data, analytics, AI, sustainability and industry research.
“From business intelligence to fraud prevention, AI offers vast opportunities – but to unlock its full potential, organisations must prioritise data integrity, governance, and innovation,” he said.
Cloud migration is still a top priority for Asia-Pacific companies, with 73 per cent saying that it is key to strengthening their GenAI capabilities. The region’s AI investment is on the growth path, with Asean’s spending expected to see an annual 28 per cent growth.
Several countries in the region are advancing the development of large language models (LLMs) that reflect local languages and cultural contexts, leveraging local data and cloud technology.
In Singapore, AI Singapore has launched Sea-Lion (Southeast Asian Languages In One Network), a comprehensive LLM for the region. In addition, Malaysian AI startup Mesolitica has created a Bahasa Melayu LLM called MaLLaM (Malaysian Large Language Model) that is hosted on the cloud.
To boost the success of building a strong data foundation, IDC advises Asia-Pacific organisations to develop enterprise-wide data strategies and promote a culture of data-driven decision-making.
They should also invest in scalable platforms and automate data management processes, while strengthening data governance and lifecycle practices for privacy and regulatory compliance.
Organisations should also enhance infrastructure, build data expertise, and pursue strategic partnerships to maximise analytics capabilities, according to IDC.
