Will artificial intelligence (AI) help smash through the technical debt that has been growing in recent years? Or will emerging technology add a new layer of issues? Or, looking at it another way, could the overhang of technical debt slow down efforts to adopt AI? The good news is that a recently issued report suggests AI may be an answer to finding and reducing systems that slow things down.
Also: Meet the most important AI tool that no one is talking about – but everyone needs
Technical debt is defined as shortcuts or workarounds taken to meet delivery deadlines quickly, according to tech analyst Gartner. The most recent figures available put accumulated software technical debt at more than $1.52 trillion, according to an estimate from the Consortium for Information and Software Quality.
Many companies have been case-hardened within the concrete of technical debt for decades. The new report and survey of 800 executives from HFS, a technology research firm, co-produced with Publicis Sapient, suggests that AI may finally be the “jackhammer” that breaks through accumulated debt.
Also: Have a genius business idea? These 2 AI tools can help you turn it into a reality
The study’s authors were inspired to conduct the research after a complicated engagement to wean a large insurance carrier off its mainframes. This encounter led them to ask, “Are we approaching technical debt the right way,” said Sheldon Monteiro, executive VP at Publicis Sapient, at a recent HFS conference. “Everybody has these tools, yet there’s this inertia.”
He said entrenched incentive plans can be blamed to a large degree for this inertia in handling technical debt. Many efforts to reduce technical debt are about swapping out older systems for new systems, “code base to code base, lift and shift to cloud.”
<!–>
The best way to tackle and potentially reduce technical debt is to recognize such efforts as a “transformational process. It’s not about doing things faster, it’s about doing things in a whole new way. It’s not an engineering problem, it’s a design problem.”
The HFS study found that 80% of executives believe AI will finally move the modernization needle. Close to three in four (74%) aim to boost their business requirements through a services-as-software approach, defined by the study’s authors as “highly autonomous, Al-driven service delivery, outcomes-based or subscription pricing.” Another 49% intend to boost their use of Al-led agentic services.
Also: How to use ChatGPT freely without giving up your privacy – with one simple trick
There’s still a lot of work to be done, though. More than half of the executives surveyed (55%) said they lack the talent, data, or governance capabilities they need to build an AI-driven enterprise. Ethics and compliance concerns were a priority for 39% of firms, and 41% are still struggling to integrate AI with legacy systems.
“Don’t manage tech debt – demolish it,” the study’s authors said. “Treat tech debt like financial debt – track it, prioritize it, and pay it down with discipline. Stop masking the problem with automation duct tape.”
Also: 5 ways to turn AI’s time-saving magic into your productivity superpower
The report suggested this area is where AI can help. “Use AI to understand, refactor, and retire legacy systems, starting with what slows your business the most.”
Then, “rebuild around AI – not on top of it: AI isn’t an overlay; it’s a re-architecture. Enterprises stuck layering AI onto brittle legacy systems will simply accelerate dysfunction. Rethink workflows, data models, and governance from the ground up, with intelligence as the foundation. where decisions are automated, outcomes are measured, and systems are designed to learn and adapt.”
