FROM THE 2026 ANNUAL REPORT · SECTOR SUMMARY
What the Asset SERVICERS sector shows
Servicers cluster at an evolving posture on the model layer at the typical level, with the strongest firms reaching evidenced control in how models are operated and the rest disclosing named tools without a corresponding control mechanism. The sector illustrates a distinction this report draws throughout: a named platform or deployed application is a capability, not in itself a control mechanism, and the tier turns on whether an auditable governance mechanism is disclosed around it. Agentic operation across the category is disclosed at an evolving stage at most — frequently in development or pilot — rather than in governed production.
Heat-map classifications reflect publicly available information reviewed under the methodology described in this report. They are not assessments or certifications of any institution’s actual internal AI capabilities or controls. Grey (Not Disclosed) indicates the absence of public disclosure, not the absence of control.
The report does not evaluate, rate, certify, or benchmark any individual institution; the tiers reflect the completeness of public disclosure as our review found it, not an assessment of any institution's actual controls.
AI CONTROL FOR ASSET SERVICERS
The infrastructure layer of institutional finance. Your clients' governance starts with yours.
For asset servicers, the challenge is not just processing transactions accurately — it is governing AI that now operates inside every workflow, under zero-tolerance SLAs, across every regulatory boundary clients impose. Your AI control posture is a link in every client's fiduciary chain.
THE ASSET SERVICING AI CONTROL CHALLENGE
Custodians & Global Custody Banks
Their Mandate: Safeguard client assets, settle transactions, and produce accurate records across multiple markets, currencies, and regulatory jurisdictions simultaneously.
Core Challenges:
- Client governance cascade → Institutional clients — pension funds, sovereign wealth funds, asset managers — are beginning to impose AI governance requirements on their custodians. Your AI posture must satisfy not just your own regulators but the governance standards of your most demanding clients.
- Data residency at operational scale → Custody operations process client data across dozens of jurisdictions. AI models processing that data create cross-border transfer exposures that standard data center governance does not address — and that DORA and EU AI Act obligations make increasingly difficult to ignore.
- Operational resilience under DORA → The Digital Operational Resilience Act creates explicit obligations around ICT third-party risk management. AI infrastructure providers are ICT third parties. Most custody bank AI vendor agreements do not satisfy DORA's contractual requirements for audit rights, exit provisions, and concentration risk management.
Fund Administrators
Their Mandate: Calculate NAVs, process investor transactions, maintain fund records, and produce regulatory filings with accuracy that cannot be questioned.
Core Challenges:
- NAV accuracy liability → AI-assisted NAV calculation and pricing validation creates a new category of operational risk. Errors due to model drift or data quality failures that go undetected create liability that flows directly to the administrator — and through the administrator to the fund's investors and directors.
- Regulatory reporting defensibility → AI-generated regulatory filings — AIFMD, UCITS, Form PF, FATCA — carry identical accuracy obligations to manually produced ones. The audit trail requirements for AI-assisted regulatory submissions are the same as for human-prepared ones. Most fund administrator AI systems cannot produce those trails on demand.
- Client data sovereignty → Fund administrators aggregate investor data — KYC records, subscription documents, beneficial ownership information — across multiple fund structures and jurisdictions. AI processing of that data under standard vendor terms creates data residency and privacy obligations that most administrator agreements do not address.
Transfer Agents
Their Mandate: Maintain shareholder records, process subscription and redemption transactions, manage investor communications, and administer distributions with complete accuracy.
Core Challenges:
- Investor data protection → Transfer agents hold some of the most sensitive investor identification data in the financial system — passport copies, tax residency declarations, beneficial ownership structures. AI systems processing this data under standard API terms are processing it on third-party infrastructure without the technical controls that investor data sensitivity requires.
- AML/KYC AI governance → AI-driven AML screening and KYC refresh workflows create regulatory obligations around model explainability, decision audit trails, and adverse action documentation. The governance framework around those AI systems must satisfy the same standard as the compliance function they are performing.
- T+1 settlement pressure → Compressed settlement cycles create operational urgency that is pushing AI automation into transfer agency workflows faster than governance frameworks are being built around them. Speed of automation without governance is operational risk accumulating silently.
Securities Services Providers
Their Mandate: Deliver integrated post-trade services — custody, fund administration, collateral management, securities lending — across complex multi-asset, multi-jurisdiction client relationships.
Core Challenges:
- Concentration risk across the AI stack → Large securities services firms often rely on one or two major cloud providers for AI workloads spanning multiple service lines and client types. A single provider disruption or contractual restriction creates simultaneous operational exposure across custody, fund administration, and reporting functions .
- Cross-service data governance → AI systems that access data across service lines — using custody data to inform fund administration, using transfer agency data to inform collateral management — create data governance complexity that no single vendor agreement was designed to address.
- Client due diligence escalation → Sophisticated institutional clients are beginning to include AI governance in their annual service provider reviews. Securities services firms that cannot demonstrate matrix-level governance across their AI stack are entering a competitive disadvantage that will compound as AI governance due diligence becomes standard.
ASSET SERVICERS — AI USE CASES
OPERATIONAL RISK & RESILIENCE
Control the operation before it controls you
Use Cases
- Real-time operational risk monitoring across all processing workflows
- AI-driven SLA prediction and breach prevention
- Cyber threat detection across client data environments
- Business continuity scenario modeling and stress testing
Value Creation
- Proactive risk mitigation before client impact
- Reduced operational losses from processing failures
- Institutional-grade resilience posture for regulatory and client scrutiny
Tie to Stack
- OLTAIX™ → policy enforcement, anomaly detection, and real-time intervention
- Models → predictive risk intelligence across operational workflows
DATA GOVERNANCE & LINEAGE
Know where every number came from
Use Cases
- End-to-end data lineage tracking across systems and counterparties
- Master data management and golden record construction
- Data quality monitoring with AI anomaly detection
- Jurisdictional data residency enforcement across client mandates
Value Creation
- Single source of truth across the servicing operation
- Demonstrable data sovereignty for regulatory and client purposes
- Elimination of data provenance gaps in audit scenarios
Governance Reality CheckData lineage is the foundation of every regulatory audit, every client dispute, and every operational investigation. AI systems that process data without producing institution-controlled lineage records create gaps that cannot be closed retrospectively.
Tie to Stack
- Data Centers + Models → governed data fabric with complete lineage
- OLTAIX™ → real-time lineage visibility and residency enforcement
RECONCILIATION & EXCEPTION MANAGEMENT
From manual resolution → governed automation
Use Cases
- AI-driven multi-custodian reconciliation across positions, cash, and corporate actions
- Automated exception identification, classification, and resolution routing
- Real-time break detection with root cause intelligence
- Cross-counterparty discrepancy analysis and escalation
Value Creation
- Dramatic reduction in manual reconciliation effort
- Faster break resolution cycles
- Reduced operational risk and client SLA exposure
Governance Reality CheckReconciliation agents that autonomously identify, classify, and route exceptions are performing operational functions that carry institutional liability. Every agent action must be traceable to a defensible, auditable record. Most asset servicer AI deployments route exceptions through vendor systems — the audit trail lives in someone else's infrastructure.
Tie to Stack
- Agentic Applications → autonomous reconciliation agents operating under fiduciary logic
- OLTAIX™ → governs every agent action with complete audit trail in institution-controlled systems
CORPORATE ACTIONS PROCESSING
Zero tolerance for error, at any volume
Use Cases
- Automated corporate action event detection and classification
- AI-driven entitlement calculation across complex securities
- Deadline monitoring and instruction routing
- Client notification and election management automation
Value Creation
- Elimination of manual processing risk on time-critical events
- Full audit trail on every entitlement decision
- Scalability across high-volume periods without operational strain
Governance Reality CheckCorporate action errors have direct financial consequences for clients and create liability that flows back to the servicer. AI entitlement calculations must be explainable and auditable to the same standard as manual ones. They almost never are.
Tie to Stack
- Models → event classification and entitlement intelligence
- Agentic Applications + OLTAIX™ → governed execution with explainable, institution-controlled outputs
REGULATORY REPORTING & COMPLIANCE
Turn reporting from a burden into a control advantage
Use Cases
- Automated generation of regulatory reports — CFTC, SEC, EMIR, MiFID II, DORA
- Real-time compliance monitoring across client portfolios
- AI-driven audit trail construction for regulatory examinations
- Exception-based compliance alerting and escalation
Value Creation
- Reduced cost of compliance operations
- Faster, more accurate regulatory submissions
- Defensible audit trails on demand
Industry SignalDORA's full enforcement creates explicit AI governance obligations for EU-facing asset servicers. ICT third-party risk management requirements extend to AI infrastructure providers. Most AI vendor agreements in use today do not satisfy DORA's contractual standards for audit rights, exit provisions, and concentration risk disclosure.
Tie to Stack
- Models + Data Centers → governed data ingestion and compliance intelligence
- OLTAIX™ → policy enforcement and continuous regulatory alignment
CLIENT REPORTING & TRANSPARENCY
Reporting that explains itself
Use Cases
- AI-generated client reports with natural language explanation
- Real-time performance attribution and portfolio analytics
- Customized reporting across client tiers and mandates
- Data quality monitoring and anomaly detection in client outputs
Value Creation
- Stronger client confidence through explainable outputs
- Reduced manual report production effort
- Differentiated service quality through real-time transparency
Industry Signal
Large institutional clients are beginning to require that service providers demonstrate AI governance across every client-facing output. The first asset servicer to offer real-time AI governance transparency as a standard service feature will set the standard others must match.
Tie to Stack
- Agentic Applications → automated report generation under governance
- OLTAIX™ → ensures data integrity and consistency across all client outputs
REQUEST The State of AI Control in Institutional Finance
The State of AI Control in Institutional Finance is available on a relationship basis to qualifying
All discussions covered under NDA. Tiers reflect public-disclosure completeness, not assessments of any institution's actual controls.
AI is a given. Control is not.™
