Working smarter and more personally in financial services.
Why now?
You know the picture: applications piling up, a lender asking for yet another document, and advisers spending more time on admin than on advice. Things are moving fast, clients expect more, and you want to stay ahead — without handing over your data and IP. Not adopting AI is not an option; a flawed implementation can be fatal. The question is not whether, but when you prepare your organisation for the future. This document outlines the developments, the processes that lend themselves to it, and how you discover it risk-free in a contained pilot.
Market developments
- AI is already mainstream among the large players. In early 2025 nearly 80% of large and mid-size insurers had AI in regular business processes (versus 21% of smaller ones), and 6 of the 10 largest Dutch banks use AI models in daily work (DNB). Those who lag behind stand out.
- Clients expect fast, digital service. Real-time insight and direct online handling are the norm; Dutch insurers rank among the world's best in digital customer journeys (75.3% vs. 59.2% globally; Deloitte Digital Insurance Maturity 2025).
Key challenges
The core challenge is clear: improve your service with AI without losing control.
- Scarcity of specialist (compliance/KYC) capacity, while routine work eats up time.
- Adopting AI without handing over your data, IP or control.
- A growing stack of compliance and audit requirements (see below).
- DORA — in force since 17 January 2025 (digital operational resilience).
- Accessibility (EAA) — since 28 June 2025, new rules for digital services to consumers.
- AI Act — phased: prohibited practices (Feb 2025) and general-purpose AI models (Aug 2025) already in force; requirements for high-risk AI in financial applications (e.g. credit scoring, insurance pricing) expected to be postponed to 2 December 2027.
- AMLR (EU anti-money-laundering regulation) — directly applicable from 10 July 2027, including an EU-wide cash limit of € 10,000; supervision via AMLA and nationally e.g. AFM and DNB.
Common workflows
The same administrative patterns recur across the sector:
MMC works modularly. We don't support a single workflow, but build out a growing set of processes step by step — at your own pace, always with human control.
Where AI supports
Within a process, AI supports the repetitive steps while your people decide:
Customer request
Request received
Document analysis
Understand info
AI classification
Prioritise
Risk assessment
Assess impact
Compliance check
Policy & rules
Human approval
Judgement
Execution
Update systems
The MMC approach: the pilot
We work modularly, step by step. Together we take one process; you invest minimally, we do the heavy lifting:
- Together — we pick one concrete process.
- You — an intake of about an hour, plus limited additional information.
- We — analyse your process and build a working version.
- Together — we test on your own data and discuss the outcome.
- You — decide whether and how to proceed. No obligation.
Is your organisation a fit?
Recognition is often a good gauge. This fits if you deal with:
- many recurring administrative processes;
- working with documents and data from multiple sources;
- human checks and compliance requirements;
- a lot of email traffic and long turnaround times;
- a shortage of specialist capacity.
Start a pilot
Start a pilot →
Sources: DNB — AI at insurers & banks · Deloitte / banken.nl — Digital Insurance Maturity 2025 · CBS — sickness absence by industry · banken.nl — KYC/CDD shortage · DORA (EUR-Lex) · EAA (EUR-Lex) · AI Act (European Commission) · AMLR (EUR-Lex). Regulatory dates may change; verified June 2026.