AI Is Rewriting the Economics of Outsourcing — Where Does That Leave Fiji?

Fiji is building digital infrastructure to attract more business process outsourcing work. Google is spending US$250 million on a data center and cable landing station at Natadola. The government has set a target of 40,000 new ICT jobs by 2030. The BPO sector — already 8,000 jobs and US$130 million in annual foreign exchange — is explicitly central to that ambition.

At the same time, generative AI is automating the exact category of work that made BPO profitable in the first place.

That tension is worth examining honestly — not to dismiss the infrastructure investment (which I covered in detail in Why Google Picked a Fiji Golf Resort Beach to Build the Pacific's Newest Data Center), but because the economic case for BPO expansion and the AI disruption story are developing on parallel tracks, and the timing matters.

What AI Is Actually Doing to the BPO Industry

The global call center outsourcing market sits at roughly $380 billion in 2026, still growing at around 9% annually. This is not a story of an industry collapsing. But it is a story of an industry changing shape faster than most forecasts anticipated.

Conversational AI tools — large language model-powered chatbots, robotic process automation (RPA), AI-driven ticketing systems — are increasingly handling the tasks that outsourcing hubs were originally built around: data entry, form processing, basic customer inquiries, first-tier support calls. These are high-volume, rules-based, repeatable workflows. They are also the most automatable category of work that exists.

The numbers quantify the direction, if not the exact pace. Industry analysts project that AI deployments in contact centers could reduce agent labor costs by around $80 billion by 2026 globally. Across India and the Philippines — the two largest BPO markets — somewhere between 2 and 3 million jobs face structural disruption this decade, with roughly 1 million directly impacted by 2030. That disruption is expected to concentrate most heavily in the 2026–2030 window.

There's a more subtle effect too. Research published via NBER found that LLM access increased call-center agent productivity by around 14% — meaning the same headcount can handle meaningfully more volume. Even where jobs aren't directly eliminated, net new hiring slows: you need fewer people to serve the same number of customers.

None of this is a sudden collapse. India and the Philippines both added BPO jobs in 2025 (roughly 120,000 and 80,000 respectively). The disruption is uneven and gradual. The established hubs are responding by upskilling — repositioning agents as AI-tool operators, quality reviewers, data annotators, and supervisors of automated workflows rather than executors of them. The question is how quickly that transition completes, and what the net employment picture looks like at the other end.

Where Fiji Sits in This

Fiji's BPO sector is not insulated from this. Local industry commentary already notes that AI-driven chatbots and RPA tools are automating data entry, document processing, and routine customer service within existing Fijian BPO operations. These are exactly the task categories the global research identifies as most at risk — entry-level, high-volume, process-driven work.

The exposure isn't hypothetical. It's already happening at the task level inside companies currently operating in Fiji.

The timing problem is structural. The Natadola cable landing station and the accompanying terrestrial fibre link between Vatuwaqa and Natadola are targeted for completion in late 2026. This infrastructure took years to plan, negotiate, and build. It will deliver genuinely better connectivity — lower latency, more redundancy, more capacity — for businesses operating from Fiji. That matters for competitiveness.

But the infrastructure was designed to attract and sustain a BPO growth model that assumed continued demand for the kind of work Fiji currently competes for. If AI adoption in global BPO continues accelerating, the upgraded connectivity might arrive at precisely the moment demand for high-volume entry-level outsourcing is declining relative to demand for AI-augmented, higher-skill roles.

The Honest Analysis

I want to be clear about what is established and what is speculation here, because the research on this topic concentrates overwhelmingly on India and the Philippines. No source I found specifically modeled AI's impact on Fiji's BPO trajectory. What follows is informed analysis, not a forecast.

The risk case: Global BPO clients consolidate work toward AI-mature hubs — India and the Philippines, which are already deep into the upskilling transition, with larger talent pools, established training infrastructure, and years of AI-tool integration experience. Smaller markets like Fiji struggle to compete for the next generation of BPO demand even with improved connectivity, because the transition from routine to AI-augmented work requires institutional capacity that takes time to build.

The opportunity case: Fiji's relatively small BPO base (8,000 jobs versus millions in the established hubs) may actually work in its favour. A smaller industry can reposition around AI-augmented, niche, or boutique outsourcing work faster than a multi-million-job market can pivot. Better connectivity and lower latency make Fiji more competitive for specialised technical support, data annotation, AI quality assurance — exactly the "human in the loop" roles that sit alongside automation rather than being replaced by it.

What's genuinely unclear: Whether Fiji's current workforce and training pipeline can make that shift quickly enough. The 40,000 ICT jobs target implies significant skills development investment. Whether that investment is concentrated on the roles most likely to remain in demand — or on the roles most likely to be automated — is a policy question that available sources don't answer.

What This Means for Businesses Operating Here

From an IT perspective, the practical implication for Fijian businesses is less about the macro BPO story and more about how AI tooling is changing what technology investment looks like at the business level.

AI-powered customer service tools, document processing automation, and workflow RPA are no longer enterprise-only. They are increasingly accessible to SMEs — and the cost curve is moving in one direction. Businesses that integrate these tools effectively will handle more volume with the same headcount. Businesses that don't will find themselves competing against operations that have.

The connectivity improvement from the Google infrastructure project is real and relevant here: AI tools work better with lower latency and more reliable uptime. The infrastructure story and the AI adoption story are connected, even if the policy framing around them hasn't fully caught up.

The Bigger Picture

Fiji's BPO bet isn't wrong — the sector is real, it employs real people, and connectivity investment is genuinely necessary for it to grow. But the strategy was designed for a specific version of the BPO industry that is already evolving away from the model it was built on.

That doesn't make the Natadola infrastructure project a mistake. Infrastructure that improves connectivity is durable — it serves multiple industries and multiple use cases beyond BPO. But it does mean the job-creation projections attached to that infrastructure deserve more scrutiny than they typically receive, and the policy conversation about workforce positioning for the 2026–2030 window is more urgent than the headline numbers suggest.

The race isn't between Fiji and other Pacific nations. It's between Fiji's pace of workforce and industry adaptation and the pace of AI adoption in the global industry it's competing in. That's a harder race to measure — and a more important one to watch.

Sources

All statistics and claims in this article are drawn from the following published sources. Projections are from industry analysts and are attributed accordingly; the Fiji-specific analysis in the final sections represents the author's interpretation, not independently modelled forecasts.

AI Is Rewriting the Economics of Outsourcing — Harvard Business Review
AI, U.S. offshore curbs threaten Philippine BPO boom — Outsource Accelerator
AI threatens millions of BPO jobs in India and Philippines — Outsource Accelerator
AI vs Outsourcing: The Future of Jobs (2026–2035) — guptadeepak.com
AI vs. Outsourcing: The Future of Jobs in the US and the Developed World (2026–2035) — Security Boulevard
Philippine BPO Industry in 2026: Growth, AI, and What's Next — Creathink Solutions
AI 'is threat to low-skilled BPO roles' in Philippines — CIPD
Call Center Outsourcing in 2026: Trends, AI & Benchmarks — The Octopus Tech
AI Was Supposed to Replace Offshore Customer Service. Now It Might Cost More. — CMSWire
AI Technology and Fiji's BPO Industry — LinkedIn (Gavidi Cakobau)
Top 5 Jobs in Retail That Are Most at Risk from AI in Fiji — Nucamp
Will AI Replace HR Jobs in Fiji? — Nucamp
AI Is Disrupting a Leading Philippine Industry — and Creating an Opportunity for Canada — Asia Pacific Foundation of Canada

Working in IT across Fijian businesses and watching how AI tooling is changing what clients ask for — get in touch if you want to talk through what this means for your operation.