Legacy tech blocks AI projects across Asia Pacific

IDC has published research indicating that legacy technology is hindering AI projects across the Asia-Pacific. MongoDB commissioned the study.

Survey data from 1,400 organisations across eight markets found that 43% said their existing architecture makes it impossible to build new applications without extensive modernisation. Respondents described those systems as too rigid, costly and slow for current requirements.

The research also identified a smaller group of companies, described as leaders, that are pulling away from their peers. Those organisations generate 71% of revenue from digital products and services, compared with 23% among mainstream peers.

Data problems

At the centre of the gap is the state of companies' data and software estates. The most commonly cited software development challenge was data management and poor-quality data, named by 32% of organisations.

Next were outdated database technology that does not support AI workloads, cited by 31%, and the difficulty of embedding security into development processes without affecting speed or innovation, also cited by 31%.

The findings suggest many companies are still trying to build AI projects on systems designed for earlier generations of software. Supporting new AI initiatives was the main reason for modernising databases and applications, with 46% of organisations naming it as their top driver.

Modernisation itself remains difficult. Nine in 10 organisations surveyed reported having experienced failed modernisation initiatives, with siloed and poor-quality data identified as the main obstacle.

IDC said the issue is becoming more urgent as businesses move from pilot AI projects to production systems that require dependable data, updated infrastructure and tighter integration across applications.

"The stakes for modernization are now critical. High-quality, integrated data is the essential fuel that determines the accuracy and performance of an AI application, making modern data architecture a foundational element of any AI strategy," said Dr William Lee, Senior Research Director, Service Provider and Core Infrastructure Research, IDC Asia Pacific. "But research shows that many organizations are being held back by their existing rigid legacy architectures that do not have the flexibility and scalability to handle the high volume of unstructured data required for AI."

Revenue divide

The report links long-term technology renewal with commercial results. It found that the strongest-performing group treats modernisation as a continuing discipline rather than a one-off programme.

Among those leading organisations, 58% are running multiple programmes to reduce legacy constraints and build cloud-ready foundations for AI systems in production. The broader market, by contrast, is still contending with older estates and fragmented data.

IDC has also forecast that organisations that do not address technical debt will face a 50% higher failure rate and rising costs for AI initiatives by 2027. That adds pressure on technology leaders to tackle older systems that may have remained in place due to cost, complexity, or operational risk.

The survey covered organisations with at least 100 employees in Australia, China, Hong Kong, India, Indonesia, Singapore, South Korea and Thailand. Respondents included developers, IT decision-makers, primary decision-makers and members of decision-making units.

MongoDB said AI has increased the urgency of the issue for senior executives, arguing that the results show technical debt is now more directly tied to revenue growth and execution risk.

"AI has made technical debt an urgent board-level priority," said Thorsten Walther, Managing Director, CXO Advisory at MongoDB. "The research is clear, strategic modernisation unlocks AI opportunities and supports a significant increase in revenue. The leaders across the region are showing what's possible when organisations ditch rigid, siloed legacy systems and move to AI-ready data platforms like MongoDB."

One example

The study cited Bendigo Bank as an example of a company modernising a critical system. According to the report, the bank moved a core banking application away from legacy relational database technology to MongoDB Atlas and used AI-assisted tools to break the work into smaller releases.

The migration reduced development time for that project by up to 90% and cut the cost to one-tenth of a traditional migration, according to the report. It was carried out without outages.

IDC outlined several steps for companies seeking to improve AI readiness. They included stronger data quality and governance, modernising architectures that slow application development, building cloud-ready hybrid operating models, and investing in skills and change management.

The figures point to a widening divide in Asia Pacific between companies that have begun reshaping their underlying systems for AI and those still constrained by older platforms, siloed data and repeated modernisation setbacks.