Blog article
See all stories »

Ready for ISO 20022? Make AI part of your migration toolkit

The ISO 20022 messaging standard will enable rapid payments processing and global settlement, with interoperability across many payments infrastructures.  As we move closer to the migration deadline across various jurisdictions and regions, the technology challenges for managing the messaging format migration are becoming increasingly clear. As organisations tackle issues and opportunities provided by the format switch, the cost effectiveness and delivery of additional value-added products and services will be key. 

Looking across key technologies in the payments space, the use of Artificial Intelligence (AI) disciplines undoubtedly sits at the top of an adoption list.

Just as we saw the huge impact mobile had on the world of fintech a decade ago, AI (and especially NLP) are having and will continue to have a significant impact in global payments, especially in wholesale, cross border and high-value payments processing. AI using NLP & ML technologies can be utilized for ISO20022 and MX migration at various levels, but I’d like to highlight three broad areas where AI can play a significant part in the format migration:

1. Format translations along with auto enrichment

In the phases between co-existence and complete format migration, NLP-based techniques can be used for format translations by structuring and normalizing data that has been free-format in the past and ensuring completeness and integrity of data, to derive elements that need to be in mandatory fields.

In addition, using unstructured data to enrich and map into correct ISO 20022 and SWIFT MX structured fields can then be used for additional fee derivation, beneficiary address validation or compliance and AML checks using country codes. Having structured information will enable better management of regulatory compliance and cost-effective AML,  with less false positives.

 2. Self-Learning for future payments enrichment

In the initial phases post migration, AI can be utilized for learning from transactions that went into exceptions, so that in future they be can enriched and auto repaired, therefore increasing STP and lowering costs.

3.       Voice-assisted customer support

During the transition phase from co-existence to absolute migration, there will be a surge in queries and concerns from customers. Customer service can be automated using NLP based voice-assisted live conversations, and chatbots can be provided as an ancillary service.


These three areas are not exclusive of course. AI and NLP technologies can be used in each facet of the ISO migration process – that’s to say: pre, during, and post stages – across processing, management and compliance.

ISO20022 migration will undeniably be a resource-heavy exercise, so the adoption of the right technologies in the right areas will be a primary success factor. I believe AI technologies will be an essential part of any organization’s successful migration toolkit.


















Comments: (1)

Ketharaman Swaminathan
Ketharaman Swaminathan - GTM360 Marketing Solutions - Pune 19 March, 2020, 15:20Be the first to give this comment the thumbs up 0 likes

"As we move closer to the migration deadline"?

LOL. Didn't SWIFT just announce that we're moving farther from the migration deadline? 

Swift delays ISO 20022 cross-border payments migration

Rajiv Desai

Rajiv Desai

SVP - US Operations


Member since

12 Mar 2020


Greater New York City

Blog posts


This post is from a series of posts in the group:

SWIFT Matters

A community that concentrates on SWIFT as an important business tool in the payments and standards world.

See all

Now hiring