Banks should treat cryptocurrency custody as a governance and controls challenge, not just a new product line. The core questions are how private keys are protected, who can move assets, how transactions are monitored, and how suspicious activity is escalated. A bank that lacks clear custody controls can create more operational and compliance risk than it removes.
What custody means for a bank
For a bank, cryptocurrency custody is the operating model that decides how digital assets are held, who can authorise movement, and what controls surround signing. That makes it closer to a high-stakes control environment than a simple product launch. The bank must be able to explain where keys live, how approvals work, how exceptions are handled, and how the arrangement fits its broader risk framework.
Custody also changes the bank's responsibilities around segregation, reconciliation, auditability, and third-party oversight. If the bank outsources any part of the stack, the bank still needs enough visibility to prove that the custody model is behaving as designed.
How banks should structure custody controls
The most important design choice is whether the bank can enforce separation between key generation, key storage, transaction approval, and transaction broadcast. Strong custody models limit any single person or system from moving assets end to end. That usually means policy-based approvals, tightly governed signing paths, and evidence that every release of value is intentional and reviewable.
Key protection should be treated as a lifecycle control, not a one-time setup. Banks need clear rules for creation, rotation, backup, recovery, replacement, and retirement of key material, plus controls that prevent a recovered or backup path from becoming a hidden bypass. A useful reference point for custody due diligence is Digital Asset Custody, which covers custody models, key ceremonies, policy engines, and assurance signals that matter in practice.
Monitoring is equally important. Banks should be able to detect unusual withdrawal patterns, policy overrides, address changes, and other signs that a custodian or internal operator is being abused. When activity is suspicious, the bank needs a defined escalation path that can pause movement fast enough to matter.
What can go wrong when banks rush custody
The main failure mode is treating custody as a technology purchase while leaving governance unresolved. If approval authority is unclear, a weak operator path can become the easiest route to loss. If key control is fragmented across teams or vendors, the bank may have no reliable answer to who could actually move funds at a critical moment.
Custody also concentrates operational and compliance risk. A single mistake in key handling, address allowlisting, exception approval, or vendor oversight can create losses that are difficult to reverse. For banks that are still building their digital-asset programme, the right starting point is often a broader control baseline, such as Banking Cybersecurity Programme Guide, so custody is governed inside one incident, third-party, and control model rather than as an isolated initiative.
That risk is why banks should also assume that custody failures can trigger both operational disruption and reporting obligations. If the custody chain cannot prove who approved what, or if assets are exposed through a compromised signing path, the bank may need to treat the event as both a security incident and a governance failure.
Risk and Threat Considerations
Cryptocurrency custody creates a direct theft path because the signing authority is itself the asset control point. If attackers compromise key material, approval workflows, or a privileged operator account, they may be able to move assets without needing to break the blockchain itself. The risk is highest where custody relies on long-lived secrets, weak segregation of duties, or manual exceptions that are hard to monitor.
Failure mechanism: The control fails when a bank cannot reliably separate key possession from transaction authorisation, or when a vendor, operator, or backup path can sign outside the intended policy path.
Impact: A successful compromise can lead to irreversible asset loss, disputed accountability, weak incident reconstruction, and regulatory scrutiny over whether the bank had effective custody controls at all.
Standards & Framework Alignment
This section maps relevant standards and security frameworks to the operational risks and controls described in this guidance.
NIST SP 800-53 Rev 5, NIST SP 800-57 and CIS Controls v8 set the governance and control requirements practitioners need to meet.
| Framework | Control / Reference | Relevance |
|---|---|---|
| NIST SP 800-53 Rev 5 | IA-9 — Identification and Authentication (Service and Organization Users) | Crypto custody depends on machine and service authentication around signing and transfer paths. |
| IA-5 — Authenticator Management | Custody security hinges on lifecycle control of keys, tokens and other authenticators. | |
| AU-6 — Audit Review, Analysis, and Reporting | Banks need monitoring and escalation evidence for custody transactions and exceptions. | |
| Recommendation — Enforce service-to-service authentication for every custody and signing workflow. Manage custody keys and authenticators with defined issuance, rotation, backup, and revocation rules. Review custody logs for anomalous signing, approval overrides, and transfer activity. | ||
| NIST SP 800-57 | PT1 — Key Management Guidelines and Lifecycle | Custody control depends on key generation, storage, rotation, recovery, and retirement discipline. |
| Recommendation — Apply key lifecycle rules to custody keys, including generation, protection, rotation, and destruction. | ||
| CIS Controls v8 | CIS-5 — Account Management | Custody requires tight control over privileged accounts that can move assets or approve exceptions. |
| Recommendation — Restrict and review privileged custody accounts and remove standing access that can move assets. | ||
Practitioner Guidance
What to prioritise: Start by documenting the exact custody chain, from key creation through approval to settlement, and identify every point where a human or system can override policy. If that chain cannot be described cleanly, the operating model is not ready for scale.
What to verify: Confirm that the bank can demonstrate segregation of duties, transaction review, exception handling, and recovery controls with evidence, not just vendor assurances. Custody is only credible when the bank can show how it would detect and stop an abnormal transfer before funds leave control.
Practitioner takeaway: The deciding issue is not whether the bank uses a custodian, a vault, or MPC, it is whether the bank can prove bounded authority over asset movement, end-to-end visibility, and a credible response when those controls are stressed.
Related resources from NHI Mgmt Group
- Why does bank custody of cryptocurrency change the compliance model for digital asset services?
- How should financial firms apply identity verification and screening controls when they expand into crypto payments and digital asset services?
- Who is accountable when digital asset firms expand banking access and custody under evolving rules?
- Who is accountable when banks onboard modern digital asset services through third-party integrations?
Deepen Your Knowledge
Free weekly newsletter
Subscribe to the NHI & AI Identity Journal
The latest on NHI and Agentic AI security – articles, research, breaches, news and events every week.
Bonus 33% off our NHI Course when you subscribe.
Reviewed and updated by the NHIMG editorial team on November 18, 2026.
NHI Mgmt Group — the #1 independent authority on Non-Human Identity, IAM, and Agentic AI security. nhimg.org