Book a demo
Topics · AI-DLC

Payments AI-DLC DORA | Moring

DORA treats an AI agent that changes records or moves payments as an ICT-supported function, so an AI-DLC rollout at an EU payments entity must evidence ICT risk management, logging, resilience testing, third-party risk, and incident reporting. Building those obligations into the lifecycle keeps agent-accelerated development inside the operational resilience framework.

On this page
  1. Short answer
  2. How it works
  3. FAQs
  4. Related reading

Short answer

DORA treats an AI agent that changes records or moves payments as an ICT-supported function, so an AI-DLC rollout at an EU payments entity must evidence ICT risk management, logging, resilience testing, third-party risk, and incident reporting. Building those obligations into the lifecycle keeps agent-accelerated development inside the operational resilience framework.

How it works

AI-DLC, the AI-driven development lifecycle introduced by AWS, positions AI agents as the primary executors of software delivery across inception, construction, and operations, with humans supplying direction, approvals, and oversight. For payments organizations the appeal is direct: engineering teams move from intent to running software at a fraction of the previous cycle time. The same speed, however, means a flawed change can reach settlement logic, fraud controls, or ledger integrations faster than any manual review cycle can catch it.

Payments infrastructure raises the stakes above most enterprise software. A single production change can touch payment processing, settlement APIs, transaction monitoring, onboarding, or integrations with card networks and core platforms. Research on AI-driven development for financial services emphasizes that speed only holds when a governance layer constrains what agents may do: permissioned tool access, human approval at defined decision points, security review, and audit evidence collected as a by-product of the workflow rather than assembled after the fact.

The practical pattern emerging across the sources is a governance harness built around the lifecycle. Engineering and risk teams configure the rules an agent operates under, verify outputs with property-based and failure testing, and keep machine-readable records of every decision so examiners, auditors, and internal control owners can reconstruct what happened. Adopted this way, AI-DLC gives payments teams velocity without surrendering accountability.

Nothing in this requires abandoning the speed that motivates adoption. The sources describe teams shipping AI-generated code at several times previous velocity once governance is configured as part of the workflow, because reviewers spend their time on decisions rather than on reconstructing context. For payments organizations, that trade, more throughput in exchange for controls that are architectural rather than aspirational, is the argument for adopting AI-DLC deliberately.

DORA, Regulation (EU) 2022/2554, applicable since January 2025, governs the operational resilience of ICT systems at EU financial entities. The regulation never says AI; it says ICT-supported functions, and an agent that changes records or initiates payment flows is one in every sense that matters. Five obligations reach agent deployments: ICT risk management, logging and event reconstruction, resilience testing, third-party risk, and incident reporting.

Translated into an AI-DLC context, the obligations become concrete engineering requirements: identity per agent, complete action logs, a rehearsed stop drill, contract terms covering model providers as ICT third parties, and incident evidence that reconstructs without archaeology. Storage and database choices matter too, since append-only audit records, per-agent attribution on every write, and tested failover are the capabilities compliance teams ask about first.

The payments angle is unavoidable because DORA applies across twenty types of financial entities and their ICT third-party providers. A payments organization running AI-DLC should map each lifecycle stage to the resilience requirement it must satisfy, so the evidence an examiner requests already exists in the workflow rather than in a reconstruction effort after an incident.

Sources:

www.eiopa.europa.eu: https://www.eiopa.europa.eu/digital-operational-resilience-act-dora_en

www.cockroachlabs.com: https://www.cockroachlabs.com/blog/dora-database-requirements-ai-agents/

www.moring.ai: https://www.moring.ai/blogs/dora-compliance-ai-agents/

aws.amazon.com: https://aws.amazon.com/blogs/industries/how-financial-institutions-can-operationalize-ai-dlc-quickly-and-effectively/

digital.nemko.com: https://digital.nemko.com/regulations/digital-operational-resilience-act

www.linkedin.com: https://www.linkedin.com/pulse/ai-governance-banking-how-build-secure-development-isodf - federalreserve.gov: https://www.federalreserve.gov/supervisionreg/srletters/SR2602.htm - ithandbook.ffiec.gov: https://ithandbook.ffiec.gov/ - artificialintelligenceact.eu: https://artificialintelligenceact.eu/high-level-summary/ - eur-lex.europa.eu: https://eur-lex.europa.eu/eli/reg/2022/2554/oj

Frequently asked questions

How does DORA apply to an AI-DLC program at a payments company?

DORA applies because agents acting in production are ICT-supported functions at an EU payments entity, whatever the lifecycle calls them. Its five relevant obligations, ICT risk management, logging and event reconstruction, resilience testing, third-party risk, and incident reporting, each translate into an AI-DLC control: scoped permissions, complete action logs, rehearsed stop drills, provider contract terms, and reconstructable incident evidence.

What DORA evidence should AI-DLC produce automatically?

The lifecycle should emit, as a by-product of execution, per-agent attribution on every write, append-only audit records, approval histories at mandatory decision points, and test results for each change. That evidence satisfies DORA's logging and event reconstruction expectations without manual assembly, which is exactly what a payments entity needs when an incident or examination sets a short reporting clock.

Do model providers count as ICT third parties under DORA?

Where a payments entity relies on hosted models to execute lifecycle work, the provider functions as an ICT third-party service, and DORA's third-party risk obligations reach the arrangement. Practical handling means contract terms covering resilience and data handling, version tracking of hosted models, and revalidation after provider changes, all recorded in the lifecycle evidence rather than in a detached vendor file.

What is an agent stop drill under DORA?

A stop drill is a rehearsed procedure for halting agent execution and revoking agent access, practiced the way resilience testing covers other recovery paths. For a payments entity the drill confirms two things under time pressure: in-flight agent work halts cleanly, and logs preserve what the agent did beforehand, so the incident report can reconstruct events without guesswork.