The complexity tax: Are organisations caught up in the AI adoption wave?

In the digital world, every groundbreaking piece of technology is perched atop something that already exists.

For example, the Internet is built over client-server architectures; SaaS is constructed above on-premises procurement cycles; Cloud solutions are deployed over security models that were originally designed for data centres. And right now, AI is stacked above them all, over a wobbly foundation that can hardly withstand the added weight.

These patterns are suggestive of something quite simple–the issue is never really with the technology that is introduced; it is with the complexity that lies deep within. 

Complexity Tax—What it means and how it accumulates

According to BCG’s AI at Work 2026 Report, India takes the global lead in workplace AI adoption with 74% of frontline employees actively using it. 

The same trend can be observed in the mid-market segment. The recent Cost of Complexity 2026 report states that India tops the global mid-market AI integration with 36% of organisations embedding AI across multiple core operations. This is over double the 15% global average.

Organisations adopt AI expecting it to operate across critical business functions that range from customer support, HR, finance, to IT and bring them together. However, the reality is painstakingly different. 

Having resorted to different platforms to cater to their varied yet immediate business needs, today, organisations have given rise to systems that function well in isolation but struggle to work in unison. This constitutes to what is called as the complexity tax – a hidden overhead cost that businesses are forced to pay due to fragmented systems, processes, and shadow technologies.

Stacking AI over this fragmented foundation exposes the latent operational gaps and introducing errors and delays that are present in the IT environment. And for this predicament, clearly, AI isn’t to blame; the underlying complexity is.

The consequences and the solutions at hand

The complexity tax comes with significant consequences. According to CoC 2026 Report, mid-market organisations lose 25% of their AI spend because of the overhead operational costs. Additionally, 88% of Indian mid-market IT leaders feel that managing AI complexities is increasing their team's workload. 

This simply means that the damages caused are not just impacting the budgets – it is also threatening an organisation’s overall productivity. McKinsey found that nearly two-thirds of organisations have not yet begun scaling AI across the enterprise, even as 64% say AI is enabling innovation and only 39% report EBIT impact.

Luckily enough, the situation is salvageable. And here’s the kicker: the solution resides in the structure, not AI.

Here are the four important structural changes that must take place to ensure that organisations emerge successful amidst the rapid AI adoption:

• Execution and experimentation must happen together: As every new wave of technology erupts, organisations remain eager to experimenting. But experimentation should never happen at the cost of execution. In fact, they should happen hand-in-hand. Organisations must clearly define the criteria and metrics to ensure the shift from experimental pilot stages to final executive production.
• Resolve fragmentation through integration: When data, tools, and processes are fragmented, organisations inherently pave the way to complexities. One way to avoid this would be by ensuring that the inbound AI tools are well evaluated so that they can be integrated into business functions where they’re needed.
• Choose prioritisation over customisation: At the outset, customizing your tools to the rising business needs might seem like the logical choice. Yet, choosing tools that can provide a time-to-first-value commitment should be the go-to for organisations that are looking to cut the complexity costs. 
• Focus on outcomes, not capabilities: Measure the success of any AI initiative undertaken through the business value it generates, not the technological aspects of it. Organisations must keep track of the improvements achieved—such as quicker workflows, reduced manual effort, or improved customer experiences. This will maintain stakeholder confidence and ensure AI investments remain tied to real operational outcomes.

Every era ushers in with promises of a new phase of digital transformation. And every generation is lured by the possibilities it presents. But organisations must realise that real value is derived from not just the technology but also the foundations upon which it is introduced. 

As AI adoption skyrockets, organisations that fail to work on their structural challenges, and cover their complexity tax will risk losing more time and resources managing fragmented systems. They will miss out on performing tasks that rake in actual business outcomes. An organisation’s success is no longer dependent upon its ability to adopt AI but on how effectively they build the structures that are crucial to support it.

Murali Swaminathan, Chief Technology Officer, Freshworks