In the first half of 2025, UK consumers lost £629.3 million to fraud, while banks prevented a further £870 million in attempted unauthorised fraud through their security systems. AI is already a key part of the defence: 73% of financial institutions are already using AI or machine learning for fraud detection, according to Thomson Reuters.
The question is no longer whether AI has a role in fraud prevention, but how far banks can trust it to act. As fraudsters use AI to automate and scale attacks, agentic AI can give banks new ways to investigate alerts, connect signals and act faster. But greater autonomy brings its own questions around explainability, governance and human oversight.
This executive roundtable brings together senior fraud, AML, financial crime and technology leaders to explore where Agentic AI can deliver meaningful value in financial crime workflows today, and how banks can move towards greater autonomy without losing control.
Welcome and participant introductions
Discussion:
- From alerts to action: How can agents cut the investigative burden by autonomously gathering context, correlating signals and prioritising the cases that genuinely need expert attention?
- Agentic threat detection: How can agents connect entities, interrogate multiple data sources and analyse behavioural patterns to surface risks traditional approaches miss?
- Dynamic Customer Due Diligence: How can Agentic AI move CDD beyond periodic reviews towards continuous, event-driven assessment — flagging meaningful changes in risk and when further investigation is needed?
- Autonomous investigations and human-above-the-loop: Where should banks draw the line between AI execution and human judgement? How far can agents take an investigation from alert to outcome — and how do experts retain the ability to review, challenge and intervene?
- Governance for the agentic era: What transparency, auditability and guardrails do banks need to verify an agent’s decisions and maintain accountability as it takes on more of the work?
Closing remarks
(75 minutes)
Speakers:
Request an invitation
This roundtable is designed for senior figures working across banking and financial services.
Fill in this short form if you meet the above criteria and would like to participate in the discussion on 20 October. All invitation requests will be vetted and approved (or declined) by the organiser.
About Elastic
The fight against financial crime has become a real-time data challenge. Fraudsters are industrialising deception with AI. Regulators demand explainability for every decision. Traditional systems can’t keep pace.
Elastic is the unified, explainable data foundation for financial crime prevention. It unites siloed systems across fraud, AML, and cybercrime, delivering real-time detection and investigation. Open, scalable, and transparent, Elastic integrates with existing platforms — reducing false positives and enabling faster, regulator-ready decisions without costly replacement.
