As organisations in Asia Pacific (Apac) continue on their path to data and AI-driven reinvention in 2025, they need to urgently address the historical issue of technical debt, or the hidden cost of outdated or inefficient technology implemented over the years.
Tech debt has a significant impact on technology transformation efforts. It affects IT budgets, hampers the ability to innovate, and impedes the adoption of new technologies. In the US alone, technical debt costs a staggering US$2.41 trillion annually, with US$1.52 trillion needed to fix it.
According to Accenture’s latest CXO Survey nearly 30% of Apac leaders flag limitations with data or technology infrastructure as the key stumbling block in their efforts to using generative AI in their organisations. This limitation is in part created by technical debt, and organisations in this region are expected to spend nearly 30% of their IT budget on remediating technical debt over the next two years, which is 8% higher than the global average.
The hidden cost of a quick fix
Modernising outdated systems to reduce technical debt has always been challenging for a variety of reasons. As a result, organisations have taken shortcuts to build new systems on top of outdated tech infrastructure, creating new problems. The approach can be compared to using duct tape to fix a leaky pipe, which stops the drip in the short term, but eventually creates a much bigger mess.
Let’s look at cloud for instance. Businesses are spending billions embracing it, and many have adopted a "lift-and-shift" approach to accelerate migration. This essentially means they’ve moved from on-premise environments to the cloud without significant modifications to the underlying infrastructure or applications running on it. While this has enabled quick and cost-effective migration, it has also created challenges like sub-optimal performance, security and compliance risks, and increased long-term operating costs.
See also: Is your cloud more fortress or folly? Reinforcing cloud security in the digital age
Today, with the accelerating pace of AI adoption which requires that businesses establish a seamless and secure cloud and data foundation, Apac chief information officers (CIOs) don’t have the luxury to ignore technical debt.
Half of them acknowledge that tech debt significantly impacts their competitiveness and want to find ways of remediating it. They know that the new technology environment they are now building will create new demands for the future and a poor foundation will create risks for the future. Also, like a rotating credit card debt, reducing tech debt will only get harder as more time passes.
Using Gen AI to transform technical debt
See also: The next milestone for AI: Building a global trustworthy AI ecosystem
The good news is that in addition to creating a sense of urgency for fixing tech debt, AI and particularly generative AI (Gen AI) also presents organisations with a game-changing opportunity to actually solve the problem.
Take, for instance, mainframe modernisation. Upgrading mainframes has traditionally been seen as high-risk and cost-intensive, with lengthy timelines and uncertain outcomes. The fact that the underlying code has been written over time in languages not in common use today has complicated things further. Gen AI can help overcome this skills challenge by automating the generation of new code, running simulations to test its performance, and ensure compliance—all within a fraction of the time. This allows businesses to tackle legacy systems confidently, unlocking significant operational efficiency and scalability.
It can also help balance technical debt in cloud adoption. We know that many organisations in Apac transitioned workloads to the cloud without resolving underlying legacy inefficiencies. Gen AI can help by easily refactoring code for enhanced performance, automating operations for greater scalability, and strengthening security protocols.
The urgency of tackling technical debt
Tackling technical debt today is the first step toward harnessing AI in the future. This is not just about fixing old systems; it's about ensuring that organisations can scale AI effectively and efficiently – like creating a solid architectural foundation so that it can support a skyscraper in the future.
According to Accenture’s latest Tech Vision research, 96% of Apac organisations will use agentic AI in the next three years. It’s important that they create the right foundation for this today, and with generative AI they now have what they need to clear that path and accelerate their AI led transformation.
Paulo Ossamu is the senior managing director and lead technology services for Apac at Accenture
