Topic starter
10/07/2026 9:25 pm
TL;DR: At MRC Vegas 2026, more than 1,800 payments and fraud leaders focused on faster attack scaling, better decision accuracy, and lower manual review dependency as AI reshapes account takeover, payment fraud, and coordinated abuse, according to Sift. The operational challenge is no longer just detection, but adaptive trust decisions that preserve conversion while resisting automation.
Editorial analysis by NHI Mgmt Group, based on content published by Sift: “MRC Vegas 2026: What Fraud Leaders Are Prioritizing Right Now”.
By the numbers:
- More than 1,800 payments and fraud prevention leaders came together at MRC Vegas 2026.
Key questions
Q: How should trust and safety teams adapt fraud operations when AI is speeding up abuse patterns?
A: Teams should treat AI-enabled fraud as a speed and scale problem, not just a detection problem.
Q: Why do static fraud rules break down against coordinated abuse?
A: Static rules break down because coordinated abuse can vary signals faster than a rule set can be tuned.
Q: What breaks when fraud review relies only on manual decisioning?
A: Manual review becomes difficult to scale once order volumes rise, because human analysts need time to inspect each case.
Practitioner guidance
- Define trust as a runtime governance decision Map fraud, identity, and customer experience decisions to a single operating model so that risk is re-evaluated at the point of action, not only in post-event review.
- Layer behavioural and network signals Use behaviour, device, velocity, and network context together so attackers cannot rely on any single signal to bypass controls at scale.
- Reduce manual review dependence Reserve human review for ambiguous edge cases and tune decision paths so that routine trust calls are handled automatically with measurable consistency.
Bottom line: AI-driven fraud is pushing teams away from static rules and toward continuous trust evaluation across the customer journey.
Explore further
View Full Forum → | NHI Foundation Course → | Our Services → | Read the full analysis →
10/07/2026 9:47 pm
Adaptive trust signals are becoming the core control plane for fraud governance. The article shows that organisations are moving away from static rule trees because AI now changes attack speed, attack variety, and the noise floor in decisioning. That shifts the governance problem from one-time detection to continuous trust evaluation across the customer journey. For practitioners, this means the right question is no longer how many rules exist, but how quickly trust can be recalculated when conditions change.
A few things that frame the scale:
- Nearly 60% of companies reported that fraud losses were still increasing in 2025.
A question worth separating out:
Q: How should teams prepare fraud controls for peak traffic events?
A: Teams should pre-tune risk-based authentication, velocity monitoring, behavioural detection, and decision thresholds for event periods before the surge begins. Peak traffic is when legitimate noise is highest and attacker activity is most likely to blend in. The objective is to keep controls adaptive enough to preserve conversion without creating blind spots.
👉 Read our full editorial: Fraud decisioning is shifting toward AI-era trust signals in 2026