Nuvem
Beyond availability: Building cloud resilience by design
artigo
23 de out de 2025
By Kristi Cunningham, U.S. Chief Technology Officer at Kyndryl
This year’s message at Sibos 2025 in Frankfurt made one point unmistakable: payments modernization is complex. Successful execution requires a new operating model that fuses core resiliency, data discipline and auditable AI. Financial institutions adopting this approach will win on speed and safety – delivering richer ISO 20022 data, the global standard for structured financial messaging – to improve interoperability and compliance, reduce exceptions and scale in real–time.
In presentations and customer conversations, my colleagues and I shared our findings from countless engagements that modernization and AI must be complementary.
Here’s more of what we discussed about how banks and other financial entities can translate digital ambitions into measurable outcomes.
Modernize for control
Real–time payments fail where interfaces, data quality and observability are weak. Today’s heavy lift entails modernizing the “plumbing” – the compliant data flows, resilient cores and integrated fraud analytics that reduce handoffs and lower run costs. That’s how banks turn ISO 20022 from a compliance box–tick into an advantage for both operations and customers.
Scale AI where it counts
The key to modernization is how reliably and securely organizations can leverage AI across hybrid IT estates. AI-infused platforms such as Kyndryl Bridge can help organizations deploy Agentic AI with enterprise guardrails for policy enforcement, traceability and continuous monitoring. For example, in a recent program assessment with a leading financial institution, we applied Kyndryl Agentic AI Framework to accelerate service delivery – advancing agility and cyber resilience – while automating compliance with regulatory standards.
Build strategy on trust
Across the financial services industry, incident responses are slowing just as faster payment rails are raising the stakes for delay. The Sibos takeaway: unify cyber tooling, automate evidence capture and institute continuous control monitoring so “prove it” is built–in, not bolted–on. In parallel, tighten third–party risk management as more payment flows depend on cloud, APIs and external providers.
Mind the readiness paradox
Execution in real–time remains low because success depends on modernizing interfaces and data, not only front–end channels. According to the 2024 Kyndryl Readiness Report, only 5% of financial leaders polled say they can handle real–time payments today. And just 20% expect to be ready within two years. The near–term priorities are clear for 78% of leaders surveyed – real–time payments, fraud reduction, third–party risk, and audit–ready AI (underpinned by compliance) according to the 2024 Kyndryl Readiness Report .
What success looks like
Winners treat modernization and AI as a singular goal:
- Modernization: Integrate AI into existing IT and eliminate technical debt and minimize disruption.
- Composable banking: Shorten time-to-market for personalized services.
- Data observability and trust: Establish a single source of truth through visibility and governance models.
- Security and compliance: Supercharge auditing and fraud detection with AI.
- Sustainable infrastructure: Use AI to meet both performance and sustainability goals.
The bottom line: By applying agentic AI that incorporates governance, system visibility and security protocols, financial institutions can speed development of new services, reduce fraud and be ready for whatever comes next in a rapidly changing competitive and regulatory environment.
