AI Trust in Financial Services
How European Financial Institutions Will Prove Their AI is continually safe
A practitioner workshop bringing together AI deployers, regulators, and risk owners from European financial services to examine what a credible AI Act conformity package looks like in the live context of fraud detection, AML screening, and sanctions monitoring.
Financial institutions deploying AI for fraud detection, AML, and sanctions screening face a structural tension: the AI Act's conformity obligations were designed for systems that can be assessed at a point in time. But the systems that actually matter in finance are not static. They drift, retrain, and adapt and their risk profile shifts accordingly.
Article 43 conformity assessments and the technical documentation requirements under Annex IV were not written with continuous assurance in mind. Most institutions currently have no credible answer to the question: how will we demonstrate ongoing conformity after deployment?
This workshop exists to change that.
Agenda
A focused half-day session, 09:00–13:00, designed to respect participants' time while leaving room for genuine working discussion.
09:00
Welcome and context-setting
The regulatory landscape as it stands: AI Act obligations for high-risk systems in financial services, and why static conformity is no longer sufficient.
09:25
Case study: Fraud detection under the AI Act
A worked example of what Annex IV technical documentation looks like for a real-time fraud scoring model and where the gaps typically appear under supervisory review.
10:10
Case study: AML and sanctions screening
How model drift in AML systems creates a compounding conformity problem and what a continuous assurance programme needs to detect, document, and escalate.
10:55
Break
11:10
What a conformity package actually looks like in practice
Structured workshop: participants work through the components of a credible conformity package; technical documentation, human oversight logs, incident registers, and post-market monitoring plans against their own use cases.
12:15
Closing panel: governance and the regulator's view
A short discussion combining three-lines-of-defence design for agentic AI with supervisory expectations on continuous conformity. (Panel speakers to be confirmed.)
12:50
Close and structured networking
Facilitated peer exchange over coffee, with participants grouped by role and use case.with participants grouped by role and use case.
Who should attend
This is a working session. Places are limited to ensure genuine peer exchange. The workshop is designed for senior practitioners with direct accountability for AI risk, compliance, or governance in regulated financial services.
Chief Risk Officers Accountable for AI risk frameworks and supervisory relationships
Compliance Officers & MLROs Responsible for AML, sanctions, and regulatory obligation management
Operational Risk Leads Owning model risk and AI system oversight in the second line
Chief Product Officers Deploying AI into regulated product and decisioning workflows
Internal Audit Directors Designing AI audit programmes for third-line assurance
AI Governance Leads Building or maturing enterprise AI governance infrastructure
What you will leave with
A clear view of the specific obligations that apply to high risk AI systems
A working template for an AI Act conformity package, grounded in the Annex IV documentation requirements
A continuous assurance framework you can adapt for your institution's AI governance structure
A practical mapping of lines-of-defence responsibilities to agentic and adaptive AI systems