If one of your AI agents had been degrading for six weeks would you know? Most people tell me no, and that’s the problem I’m digging into.

Drifting agents can generate outputs that look fine on the outside while quietly getting worse underneath. Earlier in this series I talked about the reaction cycle as the master KPI of fraud effectiveness, and how agentic AI can make that cycle dramatically faster. This time I want to answer the question that matters once you’ve deployed those agents. How do you know they’re still working.

Most dashboards answer the wrong questions and only answer whether an agent is running. AI agent monitoring means tracking an agent deliberately, and I will walk you through exactly how to do it.

• Why a degrading agent is genuinely more dangerous than no agent at all.
• How to measure fraud reaction cycle speed at each individual stage rather than just watching one lagging number.
• Why AI agent performance metrics fraud teams should track don’t need to be perfectly automated to be useful.
• The difference between human-in-the-loop agent monitoring and autonomous agent monitoring for agents making decisions at scale.
• What a rising rejection rate actually tells you.
• How to catch a silently failing autonomous agent before real damage compounds.
• A practical three-layer AI agent monitoring dashboard fraud teams can build.

• Are running any agentic fraud ops monitoring program and want a real framework for catching a degrading agent before it shows up in your losses.
• Are responsible for AI agent governance fraud policies and need language that connects technical monitoring to leadership reporting.
• Have deployed human-in-the-loop tools like investigation copilots or rule recommendation agents and want to know what to actually track.
• Are running autonomous agents, like auto-labeling or alert clustering, with no human reviewing every decision, and worry about silent failure.
• Want to build a genuine business case for AI agent ROI fraud investment using the reaction cycle instead of just automation hours saved.

If you’ve been following fraud on LinkedIn for any real stretch of time, you are probably familiar with Brian Davis and have been reading his posts. He has been the first fraud hire at many companies across physical goods, e-commerce, marketplaces, and fintech. Today he sits at the center of it all running deep dive retreats through Safeguard.

I have talked with Brian before about AI adoption for fraud and risk teams, and he said something that stuck with me. There’s a real difference between AI activation and AI enablement. Most organizations think they’ve done the second when really they’ve only done the first. I wanted to bring our conversation to all of you, because I think fraud teams may underestimate or overestimate where they sit on this journey.

• Why AI activation vs AI enablement is the distinction most companies get wrong, and what it actually looks like when you throw a tool over the fence with no guidance.
• Brian's four pillars for real AI enablement are clear policies, actual training, dedicated incentives and time, and a feedback loop that doesn't die after three weeks.
• Why change management for fraud teams is really a people management problem wearing a technology costume.
• Brian's full five-stage framework, covering AI blocked, AI aware, AI enabled, AI first, and AI native, and how to honestly assess where your own team sits.
• Why AI enablement leadership buy-in has to start at the top, and how executives showing their own AI usage removes the imposter syndrome holding everyone else back.
• Why so many fraud teams try to go big on day one, transaction monitoring, KYC, and end up frustrated, when the smarter path is workflow design for AI adoption that starts small.
• Concrete, non-technical AI use cases for fraud, including pattern analysis, internal reporting, and OKR alignment with sales and marketing.
• How reducing engineering dependency with AI is changing what fraud analyst upskilling with AI actually looks like day to day.
• Brian's personal framework for fraud practitioner AI use cases, including his own daily habits and how he thinks about build versus buy AI fraud tools.

• Are a fraud or risk leader trying to figure out whether your team is actually AI enabled or just AI activated
• Are responsible for fraud team AI training or building out fraud team AI governance policies from scratch
• Are a fraud analyst wondering how to build AI literacy without waiting for your company to hand you a roadmap
• Are trying to motivate a team through fraud team change management without losing the people who are cautious about the change
• Are comparing AI first fraud organizations against the fully rebuilt AI native fraud teams and wondering which one is actually the realistic goal

Risk leaders are under pressure right now to use AI to cut costs. Cutting costs usually means cutting headcount. In fraud operations specifically, that instinct creates a blind spot in teams.

The previous episodes in this series walked through what a transformation actually looks like moving from manual fraud operations to AI powered ones. What those episodes didn’t get into is why scaling fraud analytics has to happen alongside that shift. There’s a second order effect almost nobody plans for. One function doesn’t shrink when a team adopts AI, it has to grow. If a fraud team headcount planning doesn’t account for that, the result is a smaller team that isn’t actually equipped to govern the automated systems it just deployed.

That is fraud analytics. Skipping its growth is how AI rollout quietly turns into a bigger risk than the manual process it replaced.

• Why scaling fraud analytics matters more than any other staffing decision in an AI transformation, and most teams get this backwards.
• Why most fraud teams break down into the four functions of fraud ops, fraud analytics, fraud strategy, and data science in fraud teams. And why almost none of them have all four fully staffed.
• Why fraud ops vs fraud analytics respond in opposite directions to AI adoption.
• The is a real difference between reviewing an individual agent decision and governing a fully automated pipeline at scale.
• What silent pipeline failure actually looks like in practice, and why automated systems don’t announce when they’ve gone wrong.
• Why rule writing automation still requires human review, and what that review has to catch.
• How KPI monitoring for automated systems and root cause analysis in fraud systems are skills fraud teams already have, just aimed at a new target.
• Where fraud team restructuring for AI usually breaks down in quarterly reviews. Why it happens when missing error thresholds, and because audits happen monthly instead of weekly.
• Why fraud analysts, not investigators or engineers, are becoming the new AI team leaders.
• How to think about fraud team budget planning during this shift, including funding analytics growth from fraud ops savings.

• Lead a fraud team currently planning or mid-way through an AI transformation and haven't yet mapped what happens to your analytics function
• Are under pressure to cut fraud team headcount and need a clear argument for where that logic breaks down
• Have deployed or are about to deploy agentic AI for investigations, labeling, or rule writing and want to understand the governance gap most teams miss
• Are building a fraud team budget case for your board and need language that connects cost savings to where they should actually be reinvested
• Want a practical framework for fraud team org design that accounts for pipeline-level monitoring, not just individual case review
• Are wondering whether your fraud analytics function is sized for the automation you're already running, or the automation you're about to add

I have been wanting to have this conversation for a while.

Dajana Gajic-Fisic has 26 years in merchant fraud. She started at Macy's in 2000, calling Visa and Mastercard in multiple languages to manually verify addresses at the point of sale. She has watched e-commerce fraud detection evolve from its earliest form through chip and pin migration, the explosion of online fraud, and now the AI era. She is currently VP of Fraud Strategy at The Wolfe Companies, working in the gift card space, which if you think has no fraud, you would be wrong.

What I liked most about this conversation is that Dajana is not someone who talks about merchant fraud from the sidelines. She fights it every day, including building out merchant fraud ring detection processes from scratch and figuring out how to detect merchant fraud attacks before they ever reach a payment page. That gives her a perspective on what is actually happening right now versus what the industry tends to talk about.

We covered a lot of ground. One thing I want to flag before you dive in. This is one of those conversations where we keep coming back to basics. Not because the threat landscape is simple. But because getting the basics right is actually how you handle whatever the threat landscape throws at you next. I thought that was worth saying up front.

• Why e-commerce fraud trends follow predictable patterns when the environment shifts, and what the chip and pin migration of 2015 tells us about AI today
• How Dajana thinks about AI powered fraud attacks as a practitioner who is actively fighting them, not just theorizing about them
• Why first party fraud and refund abuse in ecommerce may be the number one threat for merchants right now, not AI
• How fraud as a service has made it possible to commit refund fraud without any technical knowledge, and what that means for merchant fraud teams
• Why nearly fifty years of chargeback dispute rules regulation have not kept pace with the environment merchants are actually operating in
• The cross-merchant fraud intelligence sharing story that could have prevented six months of losses at another merchant's business
• Why end to end fraud monitoring is one of the most neglected basics in merchant fraud prevention strategies
• Dajana's 360 approach to fraud operations, which separates the fraud process into four parts and maps how they feed each other
• Why merchant fraud KPIs like chargeback rate alone tell an incomplete story and what to measure alongside them
• How the lines between merchant fraud vs bank fraud are blurring at the identity and behavior layer

• Work in e-commerce fraud detection and want a practitioner's view on what is actually changing versus what is being overhyped
• Are dealing with first party fraud, friendly fraud chargebacks, or refund abuse and want to hear how someone with 26 years of merchant fraud experience thinks about it
• Have felt frustrated that merchant fraud collaboration and intelligence sharing stops at your immediate network
• Are building or restructuring your fraud team structure and want a framework that actually scales
• Lead a merchant fraud team and are trying to figure out how to get upstream of the payment rather than catching fraud at checkout
• Want to understand how merchant cyber security collaboration is evolving and what convergence actually looks like in practice on the merchant side
• Are newer to the space and want a fraud fighter career development perspective from someone who grew up inside the industry

Most fraud teams that have started adopting AI agents in fraud operations started in the right place. They piloted inside investigation. They enriched alerts, structured cases, and recommended resolutions for investigators to validate. And for the most part, they are seeing real efficiency gains.

The problem is that almost nobody goes further.

In part one of this series, I made the argument that the master KPI in fraud is not precision or accuracy. It is the reaction cycle. The time it takes your system to detect a gap, whether that is a new fraud attack or a misbehaving control, and ship a fix for it. Automating your investigation process is the first step in that journey. It is not the journey.

Even if you automate investigations completely, the rest of your links in the chain are still running at human speed. The rules you deploy to flag events are still degrading. The labels feeding your models are still arriving weeks late. You are running faster investigations inside a broken loop.

This episode is about closing that loop. All five steps of it.

Before we get into the notes, if you landed here first, I'd recommend going back and listening to part one. We covered quite a lot that will make this one easier to follow. Link is below.

• Why automation of fraud investigation is only the first step in adopting AI agents in fraud, not the destination
• How the fraud reaction cycle breaks down into five distinct steps, each producing the input the next one needs
• Why fraud alert clustering is the step that turns a pile of unrelated alerts into a curated set of assembled ring-level cases
• Why automated fraud labeling is the single most important bottleneck in the entire reaction cycle and how to close it
• How continuous fraud labeling at scale changes what your models and rules can do
• Why fraud risk segmentation is the most underestimated layer in fraud strategy and why it has to come before detection automation
• How the fraud rule recommendation engine in step five only works if the four steps before it are already in place
• Why fraud ops AI transformation is not a one-quarter project and what teams further along actually look like
• The organizational and governance capabilities your team needs to build at each stage before the next stage makes sense
• Why the goal is not just lower cost but a fundamentally different and better fraud organization

• Are working through the question of where to start when adopting AI agents in fraud and want a concrete, sequenced answer
• Have already deployed agents inside investigation and are trying to figure out what comes next
• Manage fraud analytics, rule writing, or model governance and want to understand where agentic AI fits into your work specifically
• Are responsible for fraud ops reaction time and want to understand how to measure and improve it at scale
• Have felt the pain of delayed chargebacks slowing down model retraining and want to understand how automated fraud labeling solves it
• Are building the business case for agentic AI in fraud and need a framework that goes beyond efficiency gains in investigations
• Lead a fraud team that is feeling pressure to adopt AI quickly and want clarity on how to do it without creating the governance problems that cause these projects to fail

Okay, this episode made me rethink something I've been hearing more and more over the past year: "Our fraud stack is AI-powered, so we're ready for AI fraud."

Honestly, that's a comforting story. It just isn't necessarily true.

Because the problem isn't simply that fraudsters have AI now. It's that AI-generated fraud is becoming increasingly difficult to distinguish from legitimate customer behavior. The old tells are disappearing, and many of the assumptions we've relied on for years are starting to break down.

In this episode, I sit down with fraud strategy advisor David Liu to discuss what undetectable AI-powered fraud actually looks like, why traditional fraud detection systems are struggling to keep pace, and how fraud leaders should rethink their fraud prevention strategy before today's attacks become tomorrow's baseline.

This conversation explores everything from deepfake fraud and synthetic identity fraud to autonomous fraud attacks, AI-powered cyber fraud, and the growing role of AI fraud detection in modern fraud operations. More importantly, we discuss how organizations can build fraud prevention systems that continue learning as attackers evolve.

• Why undetectable AI-powered fraud represents a fundamental shift in fraud risk management
• How AI-generated fraud is changing the effectiveness of traditional fraud detection models
• Why deepfake fraud and synthetic identity fraud continue to become more convincing
• How AI fraud detection can improve real-time fraud detection without relying solely on static rules
• Why fraud operations automation should focus on accelerating learning rather than replacing analysts

• Lead fraud operations or fraud strategy teams
• Are evaluating AI fraud detection solutions
• Want to modernize your fraud prevention system
• Need a stronger fraud prevention strategy for AI-enabled attacks
• Want to understand how autonomous fraud attacks are changing the fraud landscape

Every fraud leader I've spoken to this year has heard the same message: cut costs. You've probably already had the conversation. You've explained why fraud isn't just another operating expense, why reducing investigators today often means higher fraud losses tomorrow, and why your team is already stretched thin. And yet the budget cuts are still coming.

Here's the question I'd rather ask. What if finance isn't actually your biggest problem?

In this episode, I introduce the idea of agentic fraud ops and explain why the real issue isn't shrinking budgets. It's that most fraud organizations are still operating at human learning speed while their adversaries have already moved to machine speed.

We talk about why fraud detection systems decay over time, why fraud rules and machine learning fraud detection models become less effective the moment they're deployed, and why the future of fraud prevention AI isn't about replacing analysts. It's about building systems that continuously learn.

Honestly, optimizing a slow system is still optimizing a slow system.

This episode explores what happens when AI-powered fraud detection becomes part of the reaction cycle itself instead of just another automation project. If we're going to redesign fraud operations, we first need to rethink what performance actually means.

• Why agentic fraud ops changes how fraud teams should measure success
• Why every fraud detection system begins decaying the day it goes live
• How fraud detection rules, machine learning models, and manual review age differently
• Why reaction speed is becoming the most important fraud metric
• How AI agents for fraud detection can compress learning cycles from weeks to hours
• Why fraud operations automation should focus on new capabilities instead of replacing people
• How fraudsters are already using AI-driven fraud prevention techniques against defenders

• Lead a fraud operations or fraud strategy team
• Want to modernize your fraud prevention system
• Are evaluating AI agents for fraud detection
• Need to reduce costs without increasing fraud losses
• Are planning the future of your fraud operations transformation

So this episode starts with a data executive walking into a fraud team.

Which sounds like the setup to a very niche joke. And maybe it is. But honestly, it also describes a real problem most fraud teams know too well: we rely on data for almost everything, but the relationship between fraud teams and data teams is often messier than anyone wants to admit.

In this episode, I’m joined by Shachar Meir, a data advisor and former director of data at Meta, to talk about LLM analytics, self-service analytics, data infrastructure, and why fraud teams cannot just throw AI on top of messy data and hope it becomes strategy.

Because yes, LLM-powered self-service analytics sounds amazing. Ask a question in plain English, get an answer, move faster, avoid waiting three weeks for a data team with 900 priorities. Great. I want that world too.

But then you’ve got to ask yourself: where did the answer come from? Which tables did it join? What definition did it use? Did it understand the business context? Did it hallucinate? Did you ask the right question in the first place?

Not a small detail.

This conversation is really about the gap between the promise of AI analytics tools and the operational reality of fraud analytics. Fraud and risk teams work in adversarial environments. The problem keeps changing. The signal-to-noise ratio is bad. The cost of getting it wrong can be massive, whether that means letting fraud through or blocking legitimate users. So if you want LLM analytics to actually help, you need more than a shiny interface. You need data foundations, governance, semantic clarity, and the humility to start small.

• Why fraud and risk teams are unusually complex data customers
• How fraud teams can work better with data teams instead of requesting endless point solutions
• Why self-service analytics failed so often before LLMs entered the picture
• What changes, and what does not change, when LLM analytics becomes the interface
• Why asking the right data question matters more than getting a fast answer
• How data governance, semantic layers, and data quality shape AI analytics results
• Why fraud teams should start AI adoption with one curated table, one use case, and one measurable outcome
• How to think about the one-hour, one-day, one-week, and one-month versions of a data project

• Work in fraud operations and depend on data teams to ship risk or fraud analytics projects
• Are considering LLM analytics or AI agents for data analysis inside your fraud stack
• Have been promised AI data querying that sounds too easy, because it probably is
• Need better ways to align with data engineering, analytics, or risk data infrastructure teams
• Want a practical way to think about AI data governance before deploying tools that affect real users