AI Is Driving Up Infrastructure Costs Inside Call Centres, Not Cutting Them

The promise of artificial intelligence in the business process outsourcing (BPO) industry has centred on efficiency gains and cost reduction. But for many South African call centre operators, the reality is proving to be the opposite, as the infrastructure required to run AI workloads pushes operational costs higher.

That is the assessment of Head of GBS/BPO Solutions at Qrent, Sanjay Govender, who argues that the BPO sector has underestimated the backend demands of embedding AI into customer engagement environments.

From voice neutralisation software and real-time call assistance to AI-driven first-line support and live agent coaching, the processing requirements inside modern BPO environments have increased substantially over the past 18 months.

“The uncomfortable reality is that AI is not automatically reducing operational costs inside BPOs,” Govender warned. “In many cases, it is increasing them. The difference is that the costs are shifting away from people and moving into infrastructure.”

The Hidden Cost Of Running AI

While much of the attention around AI adoption in call centres has focused on software licensing and capability, Govender pointed to the backend compute requirements as the real cost driver.

AI workloads require compute power, memory, networking throughput, low-latency environments and infrastructure that can operate at scale. Many BPO providers, he argued, did not anticipate how quickly those demands would escalate once AI tools became embedded in day-to-day operations.

The result is that operators now face a series of expensive infrastructure decisions, each with its own trade-offs.

Option One: Upgrade The Desktops

One approach is to run AI workloads directly on endpoint devices. This means moving away from standard workstation deployments toward higher-specification machines capable of handling AI-assisted applications locally.

In practical terms, this is driving a shift away from traditional Intel i5 deployments toward growing demand for i7-powered devices on the call centre floor. AI-enhanced workloads are forcing hardware upgrades well ahead of many refresh cycles originally planned for.

Option Two: Push It To The Backend

The alternative is to keep endpoint devices relatively standard while shifting the AI processing burden into the backend environment. In this model, AI applications and workloads are hosted centrally on servers, reducing the processing demand on the user device itself.

While this avoids large-scale desktop upgrades, it introduces a different problem: the server infrastructure requirements increase considerably.

Backend server environments capable of supporting AI-driven workloads require higher compute density, increased storage performance, more advanced networking and greater scalability than traditional call centre infrastructure has typically provided.

The cost of expanding on-premises server stacks to accommodate these workloads is rising, particularly as demand for AI-capable hardware continues to grow globally.

Research from Gartner indicates that worldwide spending on AI-optimised servers is accelerating as organisations move to support enterprise AI workloads, contributing to overall global IT spending reaching US$6.15 trillion in 2026.

Option Three: Move It Off-Premises

A third route being explored by many organisations is moving AI infrastructure off-premises entirely through hyperscale providers such as Amazon Web Services or colocation environments.

In this model, the infrastructure is rented rather than owned, with AI workloads hosted externally and delivered to the BPO environment through cloud or hosted platforms.

While this removes the burden of large upfront infrastructure investment, it introduces ongoing rental and operational expenditure costs that need to be managed over time. For some BPOs, this creates greater flexibility. For others, particularly those operating at scale with strict latency and compliance requirements, the long-term cost equation becomes more complex.

From Labour Costs To Compute Costs

Govender argued that AI is fundamentally reshaping the economics of the BPO industry. For years, cost optimisation in call centres focused largely on labour efficiency. Now, infrastructure efficiency is becoming equally important.

The conversation is shifting from how many agents a BPO can support to how much compute power it takes to support them effectively in an AI-enabled environment.

This is putting pressure on the traditional procurement model. Many operators still attempt to purchase server infrastructure outright through large capital expenditure projects. But in a market where AI workloads are evolving rapidly, hardware demands are changing constantly and infrastructure pricing remains volatile, locking large amounts of capital into fixed infrastructure carries growing risk.

Leasing Over Buying

A growing number of BPOs are exploring leasing and rental models for backend AI infrastructure as an alternative to outright purchasing.

Rather than buying expensive server environments upfront, providers can deploy infrastructure through operational expenditure models that spread costs over time while maintaining flexibility as AI requirements evolve.

This approach also reduces the risk of overinvesting in hardware that may become insufficient or obsolete sooner than traditional infrastructure cycles would have allowed for. In an AI-driven environment, Govender contended, scalability and adaptability are becoming more valuable than ownership.

The New Competitive Battleground

The broader implication is that the competitive dynamics of the BPO industry may be shifting. Where the industry has historically competed on labour cost, the next differentiator could be the ability to fund and sustain AI infrastructure at scale.

“The next competitive battle in the BPO industry may not be about who has the cheapest labour model,” Govender observed. “It may be about who can afford to power AI at scale.”

Last Updated on August 27, 2026 by Nick Ross