Skip to main content
Welcome to Deloitte

If we have selected the wrong experience for you, please change it above.

Abstract digital artwork features a luminous circular ring surrounded by reflective spheres. The scene is set against a dark background with vibrant gradients of purple, blue, and orange light. Smaller glass-like orbs rest on the surface.

Reinventing the asset servicer

AI, offshoring, and the new competitive battleground in fund administration

Authors:

  • Thibault Chollet | Partner - Alternatives
  • Piotr Zatorski | Senior Manager - Alternatives
  • Gerard O’Mahony | Analyst - Alternatives
  • Most private markets asset servicers are implementing AI use cases. Few have a clear vision and strategy on where they want to be and how they intend to leverage AI
  • Pricing pressure, AI and tech-enabled new entrants, and potential re-insourcing are the main drivers of AI in the Asset Servicing market
  • AI versus offshoring is not a straightforward answer and it depends on context. What matters is a clear risk-benefit analysis of the two, done now before the next offshoring decision
  • There are two strategies when working with AI: staying relevant by doing the same work better and going beyond by offering services that were not possible before
  • Servicers that do not meet the AI baseline will fall behind. Only a few leaders will compete at the edge, where the game is capability and product rather than cost

Introduction

This article is the third in a series1, 2 on AI transformation in private markets. The first two articles examined AI from the GP perspective. This article shifts the lens to asset and fund servicers, which face the same pressure while navigating their own commercial realities. We focus on core fund administration, where the real question is not whether servicers will use AI, but how they will sequence AI and offshoring, and where they will choose to differentiate.

Where the market stands: activity without transformation

Many private markets servicers are running AI projects, yet few report live implementations with material impact. Most invest broadly but achieve limited depth, concentrating on isolated tasks such as document ingestion, data extraction, and templated reporting. Few have coordinated these tools across workflows, and manual processes still dominate operating models.

The market is not homogeneous. Tier 1 servicers, with more than EUR 100 billion in assets under administration, prioritize client-facing capabilities: digital channels, modernized fund accounting, and data infrastructure that can meet LP demands for transparency. Tier 2 servicers focus more heavily on containing costs amid fee pressure. Both face the same constraint: their data infrastructure cannot yet support their stated AI ambitions.

A change in how clients buy unites the market. AI now features routinely in RFPs and DDQs, and buyers no longer ask only whether a provider uses AI; they want evidence of outcomes such as faster onboarding, higher data quality, real-time reporting, and API-based access. Some GPs already use AI to draft their own RFP responses, raising the bar for a credible answer. Servicers without a visible roadmap will find it harder to defend mandates at renewal.

The main threats that matter

Servicers face three headline threats.

  1. Pricing pressure is the most immediate: clients expect the same service at the same or lower cost, or they will look elsewhere.
  2. Tech-enabled entrants build on modern technology without legacy constraints. This category extends beyond AI-native fintechs to software providers moving into services. Few have demonstrated sustained delivery at scale in a regulated environment, while decades of integrations and client trust remain difficult to replicate. Partnership appears more likely than displacement, so servicers should treat these firms as potential enablers as well as competitors.
  3. Automation also creates a re-insourcing risk by weakening the case for outsourcing work that technology can largely perform - and by exposing gaps in the service GPs receive today. In practice, large GPs may automate selected activities in house, while smaller GPs will more likely co-source than build regulated infrastructure. Servicers can access the same automation, however, and should adopt it before clients do.

Pricing pressure forces a cost-base decision that many servicers continue to defer. Fee compression has structural limits: administration fees often pass through to LPs, and the cost of failure in mission-critical services can outweigh the savings from switching providers. Yet the floor is not the ceiling. Every servicer must choose among AI-first transformation, further offshoring, or a deliberate sequence of both.

Extending offshoring follows a familiar playbook, with predictable savings and understood risks. But it also locks in a cost structure that AI may undercut within 24 to 36 months across a meaningful subset of workflows. Before moving more labour offshore, servicers should ask one question: can automation perform this work more cost-effectively than an offshore team? If the answer is yes—as it often is for predictable, repetitive workflows—automate rather than relocate. If the technology is not ready, use offshoring as a temporary bridge. Some work must remain local: bookkeeping can sit offshore, while fund controlling and oversight stay with the regulated entity. In either case, preparing a process for AI forces teams to rationalize and standardize it, strengthening service regardless of where the work ultimately sits.

Where AI creates the most value

Servicers should distinguish between two areas of focus because each requires a different strategy. Staying relevant means doing the same work better. Going beyond means using AI to deliver what was not realistically possible before. The first keeps a servicer in the market; the second creates competitive advantage.

The baseline

The baseline covers manual work that every servicer performs and that differentiates none: invoice reconciliation, AML screening, and document ingestion. Doing this work well is the price of staying in the market, not a way to win - and the opportunity extends beyond cost. The deliverable may remain the same NAV, but servicers can transform the experience around it through faster onboarding, cleaner data, and self-service access. Vendors will deliver much of this value by embedding AI in core systems, and servicers should press them to move faster.

The greatest baseline value comes from bridging gaps between technologies. Reconciliation across fragmented internal platforms offers one example because no single vendor has visibility across fund accounting, transfer agency, banking, and portfolio data. An agent that ingests data from each platform, normalizes it, and surfaces exceptions can shorten cycle times and free teams for higher-value work. Time tracking offers another opportunity. Servicers want to price services based on time spent but often lack reliable data because activity spans email, calendars, and messaging platforms.

The edge

The edge covers work that servicers simply could not perform before AI. Here, firms compete on capability and product by turning new possibilities into services that clients can buy.

Full-population review offers one example. Today, teams review valuations and NAVs on a sample basis because people cannot check every item. AI makes full coverage feasible by checking every accrual, manual journal, and valuer assumption against prior periods, side letters, and peer funds. This creates a different control standard, one that servicers can offer to LPs and regulators as a distinct service.

Cross-fund pattern intelligence offers another. A servicer that administers hundreds of private equity funds holds a dataset no single GP can match, spanning valuation multiples, fee structures, waterfall mechanics, and expense ratios. AI can turn that scale into benchmarking, anomaly detection, and market intelligence products that only large incumbents can offer.

A third opportunity lies in reading the previously unreadable. Side letters, LPAs, and subscription documents define how a fund works, but that knowledge often sits in the heads of a few people. An agent that can query the fund's entire legal estate in seconds can transform how teams’ access and apply that knowledge.

Continuous NAV signals provide a fourth example. AI can generate an intra-period shadow estimate that flags material movements between quarterly NAVs - a capability that LPs increasingly want but that human teams cannot deliver economically.

Servicers still need to prove the value case at the edge, making structured pilots with clear metrics the right approach. Whether pursuing edge or baseline opportunities, they should apply the same discipline: validate the use case, measure the outcome, and scale step by step.

Baseline vs the edge

Expand diagram

Three imperatives for getting it right

Define the AI ambition before deploying. Decide whether to lead the market, follow closely, or do the minimum required to remain credible. That choice should shape workforce planning, vendor conversations, the approach to offshoring, and the commercial model. Both leadership and fast followership are defensible; leaving the choice implicit is not.

Get the foundations right before scaling.

  • Start with harmonized processes. Standardize work behind the scenes while preserving the customization clients value. AI increasingly resolves this trade-off by producing client-specific outputs from standardized processes.
  • Build the data foundation next: a unified backbone on a cloud platform, master data across funds, investors, and portfolios, standardized identifiers, and APIs that replace spreadsheet bridges.
  • Then establish the agentic platform. Deploying one agent is a project; running an estate of agents across a regulated operation requires an operating model. We will explore that model in the fourth article in this series.

Embed AI in the client relationship, not just the back office. Edge capabilities create an advantage only when clients can use them. Servicers should feed validated data into client systems through APIs, surface analytics within GP workflows, and deliver new services directly rather than through stand-alone reports. This approach can shift the relationship from vendor to embedded operational partner, raise switching costs, and support capability-based pricing.

Conclusion

AI will reshape fund servicing. Servicers that deliberately sequence AI and offshoring, invest in sound data foundations, and turn new capabilities into client-facing products will be best positioned to move from commodity providers to embedded operational partners.

The real battleground is not whether Asset Servicers use AI, but how they sequence AI and offshoring, and where they choose to differentiate.

Discover our Future of Advice Blog Homepage

Did you find this useful?

Thanks for your feedback