Title

When AI breaks the BPO model: A commercial reset for financial services

Section
Insight
Summary

Authors: Jonathan Wilson and Andrew Cameron
Contributor: Thomas Karger

The Business Process Outsourcing (BPO) model in financial services is breaking down. Artificial intelligence is fundamentally upending the economics, scope, and relevance of outsourcing.

For decades, BPO contracts relied on gradual efficiency gains driven by labor arbitrage and process optimization. AI is collapsing that logic. It can eliminate entire process steps, reduce manual reviews, and compress cycle times dramatically – eroding the labor-based commercial models on which outsourcing has historically relied.

As a result, organizations must fundamentally reconsider their approach to BPO: what to outsource, how to price it, and how to structure contracts in an AI-enabled environment. In financial services organizations in particular, intense regulatory accountability raises the stakes. Banks, insurers, and asset managers must balance cost pressure and transformation ambition with governance, accountability, and operational resilience requirements that cannot simply be automated away.

Incremental contract management is not enough for financial services firms. Procurement, operations, and transformation leaders must rethink BPO at a structural level. This requires a clear reassessment of where BPO remains appropriate and a disciplined redesign of pricing, scope, and deal structures – areas where procurement leadership will be decisive.

Re-segmenting the BPO value proposition in financial services

AI is forcing a clear segmentation in the types of work organizations choose to outsource . The first step is clarity: being explicit about where each process sits in this segmentation, distinguishing between processes that require BPO and those that do not.

  1. Simple, high-volume, rules-based processes are increasingly automated in-house. Intelligent automation, AI agents, and Robotic Process Automation now allow enterprises to internalize activities that were previously outsourced purely for labor arbitrage. In these cases, BPO simply becomes redundant.
     
  2. In more complex processes, traditional labor-based outsourcing is losing credibility. When providers can deliver the same volumes with a fraction of the workforce, deals structured around full-time equivalents no longer reflect how work is actually done. Continuing with these models means risking paying for capacity that is not being used.
     
  3. At the most complex end of the spectrum, AI does not remove the need for outsourcing. Financial services involves many regulated, exception-heavy, and judgment-reliant processes such as KYC and AML remediation, third-party risk assessments, and complex claims handling. These still require human oversight, embedded controls, and accountability, meaning that full, straight-through automation is often neither feasible nor desirable.

    In these areas, the right BPO providers can become more valuable – not as sources of simple labor arbitrage, but as operators of strictly governed AI-enabled services. However, very few providers currently operate at this level, and it is unlikely that providers can make the move from low-end delivery to significantly more complex, managed services. 

Even where organizations do find providers with the right capabilities, commercial structures – including pricing and deal terms – have not kept pace with the reality of how value is now created as AI adoption escalates.