Asset managers are right to feel under pressure. Despite heavy tech investment, rising costs and the shift to low-fee products have led to costs outpacing revenue for more than a decade.
The new wave of AI—autonomous agents capable of executing complex, multi-step workflows—offers a new way to fight this problem. Early evidence suggests that a traditional asset manager with a cost of 15 to 20 basis points that reshapes their organization to deploy AI at scale could reduce expenses by 3 to 6 basis points, perhaps a 25% to 30% cut.
But the more significant opportunity is broader than just cost savings; leading asset managers are using AI to drive revenue and reduce risks. In distribution, AI acts as a force multiplier, enabling coverage and personalization at a scale no human effort alone can achieve. In investment research, analysts equipped with AI agents could monitor and conduct deep research on five times as many companies.
This is the AI-first asset manager: not an organization that has given employees access to copilots, but one that has reimagined how work gets done. Realizing the full opportunity, however, requires the CEO, CTO, and CIO to align on a common strategic vision. They should begin by selecting two or three high-value workflows and rebuilding them end-to-end around agentic AI.
AI Agents Break the Cost-AUM Link
Over the past 15 years, global AUM has more than tripled, revenues have more than doubled, yet margins remain essentially unchanged, according to BCG’s Global Asset Management Report 2026. The reason: costs have historically scaled with assets. Conventional cost-cutting does not break the structural link between rising costs and increased AUM.
The rapid improvement in AI, particularly autonomous AI agents, creates a new equation. Executing complex, multi-step workflows with minimal human intervention enables them to deliver substantial savings in tasks such as reconciliation and NAV oversight. More fundamentally, AI agents decouple costs from AUM—and reinvesting the savings keeps that curve flat even as assets grow. AUM per head can rise significantly with a limited increase in marginal costs.
AI Agents in Distribution: More Clients, Lower Cost-to-Serve
AI agents can create a step-change in the economics of distribution through:
- More effective sales to existing clients. Improved servicing and more timely outreach can capture more wallet share from today’s clients by driving higher cross-selling rates and retention.
- Expansion of the client base. AI-driven efficiency helps turn more prospects into clients. It also increases the addressable client base. A manager who previously handled only custom mandates above $500 million can now profitably offer them at significantly lower amounts.
- Unlocking capacity. Many distribution activities are ideal for AI automation. Managers have a choice: they can realize them as savings or reinvest them in more distribution activity to drive growth. The mix should be driven by firm strategy, market position, and the potential for revenue generation from strategic reinvestment.
Given this choice, impact can be highly variable. But our modeling shows that an AI-first asset manager can, within three to five years, achieve incremental annual inflows of 0.5% to 1% of AUM just through deploying AI to support distribution.
AI Agents in Investment Research and Portfolio Construction: Broader Coverage, Better Decisions
Today’s investment workflow has been digitized but remains generally unreformed. Analysts use a terminal to monitor names, diligently read earnings transcripts, issuer filings, and reports, and present to investment committees who may not have had time to prepare deeper questioning. Portfolio construction is typically weekly or monthly, and risk analysis is often siloed.
The process can be re-engineered around AI. AI agents can continuously scan thousands of names and present the analyst with curated, ranked opportunities for review. Before an investment committee meets, a committee of specialized AI agents can systematically stress-test the investment thesis—generating counterarguments, surfacing contradictory data, and flagging cognitive biases in the original analysis. AI can then help the investment committee focus on the few calls that really matter. Portfolio construction and risk analysis become real-time, quickly responding to market movements and changing correlations.
AI can increase research coverage two to five times, supporting improvements in the Sharpe ratio.
From Strategy to Execution: Building the AI-First Asset Manager
Building an AI-first asset manager requires strategic focus across seven areas, including governance and data infrastructure, as well as organizational design.
- Set ambition and focus from the top. Best results only come from CEO sponsorship with a board mandate. Budgets need to be set for at least three years and protected.
- Pick two or three workflows to start and go all-in. Concentrate on a small number of key workflows to rebuild around agentic AI and set concrete P&L targets. For most managers, this should be in research and distribution.
- Invest in activation as well as technology. Skills such as writing prompts and AI agent design need to be widespread, from the CEO to junior analysts.
- Establish the data and technology infrastructure. Only high-quality data allows agents to deliver reliable results at scale. Most managers should plan for a hybrid tech strategy; buying systems is quicker, but in-house development delivers bespoke results and avoids lock-in.
- Build governance from day one. A senior committee must ensure that auditability, human intervention, and other key issues are considered at the design stage. Retrofitting is much harder and often slows organizations down.
- Tie incentives to outcomes. If it is not in a scorecard or review, it usually will not happen. Compensation and promotion should be linked to AI adoption and outcomes.
- Plan to be the asset manager of the future. In the near term, engineers skilled at building agents will need to be embedded in the business. Long term, talent strategy must evolve to support leaner and more agile organizations.
BCG’s Build for the Future Survey 2025 shows that AI leaders among asset managers are already achieving three times the average cost and revenue benefits as AI laggards. The inherently compounding nature of AI implies this gap will widen. The opportunity to develop an AI-first strategy is open, but it will not stay open long.
