#11356 min
FRAUDFORWARD
#114SardineCon 2026 Debrief: Live from the Show Floor
33 min
What’s up fraud fighters, and welcome back to Fraud Forward!
This one’s a little different. I recorded this episode live on the floor at SardineCon 2026, and instead of one long sit down with a single guest, you're getting a debrief of the day. One conversation after another, right where they happened.
I kicked things off by catching up with Simon Taylor after his opening keynote. From there I sat down with Tal Link, Vice President at City National Bank, who’s spent 17 years in payment risk and consulting. Everything from chargeback abuse to friendly fraud to credit risk. Then came the conversation that’s going to stick with you. Joseph Cox didn’t just talk about AI powered fraud in theory. And to close out the day, Karisse Hendrick and I got into a full recap of Sardine Con.
If you couldn’t make it this year, consider this your SardineCon 2026 debrief. Straight from the people who were actually there.
What you’ll hear:
- Simon Taylor on whether the industry actually understands AI or is just moving fast enough to look like it does
- Tal Link on 17 years in payment risk, and the practical takeaways she's already applying from SardineCon sessions on AI and AB testing
- Joseph Cox walking through how he built and deployed a fake ID against a real KYC process, and what it revealed about where fraud detection still falls short
- Why agentic commerce came up in almost every conversation on the show floor, and what it means for how fraud teams need to start thinking
- Karisse Hendrick on her SardineCon roundtable, where the room debated who owns liability when an agent makes the purchase
- A rundown of the sessions, fireside chats, and roundtables that shaped the day, plus a look at Sardine's own workflow product
You should listen to this episode if you:
- Couldn’t make it to SardineCon 2026 and want the debrief straight from the floor
- Are trying to figure out where agentic commerce fits into your fraud strategy
- Want to hear what it actually looks like when someone tests fake ID detection and KYC testing against real defenses instead of just talking about it
- Are building out deepfake detection and want perspective from people who are living it day to day
- Want to hear how a traditional bank is thinking about fraud alongside the fintech and crypto crowd
- Like your fraud content with a little conference energy instead of another studio recording
Episode notes
Simon Taylor: Are we actually keeping up with AI?
The same energy from Simon’s opening speech, and his challenge to the room to find one thing they could use the very next day, carried over into our conversation. We got into whether the industry has genuinely caught up to AI or if we’re all just moving fast enough to look like we have. We also discussed what skills become more valuable once the tools get this good.
Don't confuse using AI daily with actually understanding it. Ask your team where the gaps still are.
- Don’t confuse using AI daily with actually understanding it. Ask your team where the gaps still are.
- Judgment and curiosity are becoming more valuable than technical AI skills, not less.
- Give your team the same challenge Simon gave the room and try to leave every AI conversation with one thing you can apply tomorrow.
Tal Link: 17 years of payment risk, applies
Tal walked me through what she's learned across chargeback abuse, friendly fraud, and credit risk. Then got specific about what she's taking home from SardineCon itself, from a live demo of Sardine's workflow product to a quiet return of AB testing as a fraud strategy tool.
- AB testing is coming back into fraud strategy conversations. If your team dropped it, it may be worth revisiting.
- Building and testing models on smaller data sets before scaling up saves real computing effort. Bring this back to any team building in-house models.
- Take one page from Tal's approach and leave every conference with one thing you can actually implement, not just information.
Joseph Cox: Testing AI powered fraud instead of just talking about it
Joseph didn't stop at theory. He built a fake ID, ran it through a deepfake, and successfully got it past a real crypto exchange's KYC process. We talked through exactly how he did it, what it exposed about the gaps in identity verification today, and why fake ID detection has to be stress tested against real attacks and not just modeled on paper.
- If your KYC process hasn't been tested against a real, deployed deepfake, you don't actually know if it works.
- Theoretical fraud scenarios and tested fraud scenarios are not the same thing. Prioritize the second.
- Schedule a real red team exercise against your identity verification flow, don’t just have a tabletop discussion.
Agentic commerce, the topic that followed us all day
This one came up in nearly every conversation on the show floor, not just the ones you'll hear on this episode. Karisse’s own roundtable spent 45 minutes on it, surfacing a real split. Agentic commerce for big-ticket consumer purchases still feels far off, but B2B agentic transactions, and low-friction purchases like groceries, are already closer than most people think.
- If agentic commerce isn't on your team's radar yet, it needs to be. It was the most repeated topic at the entire conference.
- Liability ownership between card brands, merchants, and AI companies is still unresolved. Start tracking how this shakes out, because it will affect your dispute process directly.
- There's no data point yet that distinguishes an agent-initiated purchase from a human one. If you're thinking about fraud prevention for agentic commerce, this is the gap to watch closing.
Karisse Hendrick: A full SardineCon 2026 debrief
Karisse and I closed out the day comparing SardineCon 2026 to last year's event. Then we went deep on her own roundtable on agentic commerce, including a lively debate about whether stablecoins or micropayments make more sense for agent-driven purchases.
- Revisit the assumption that agentic commerce means big purchases. The stronger near-term case may be small, recurring, low-friction ones.
- If you haven't asked your own team what data point would prove a purchase was agent-initiated versus human, that's a useful exercise to run this quarter.
- Cross-industry perspective matters. Karisse's most useful insights came from people outside her usual practitioner circle.
Key takeaways
- Understanding AI right now is less about mastering the tool and more about staying curious enough to keep testing it
- Fake ID detection and KYC testing need to be stress tested against real, deployed attacks, not just theoretical ones
- Agentic commerce was one of the most discussed topics on the show floor, and fraud teams that aren't thinking about it yet are already behind
- Liability for agent-initiated purchases is still an open question between card brands, merchants, and AI companies
- AB testing is making a quiet comeback as a fraud strategy tool, even outside of traditional use cases
Final takeaway
If SardineCon 2026 makes one thing clear, it is that the fraud fighters who are ahead right now aren’t the ones with the most theories about AI powered fraud. They’re the ones who went out, built the fake ID, ran the test, and found out what actually breaks.
As always, stay vigilant, stay informed, and keep moving fraud forward.
Episode resources & links:
Connect with Joseph Cox | LinkedIn
Co-founder and Journalist at 404 Media
Author of DARK WIRE: The Incredible True Story of the Largest Sting Operation Ever
Connect with Karisse Hendrick | LinkedIn
Host of the Fraudology Podcast
Award-Winning Cyberfraud Expert
Ecommerce Fraud Prevention Consultant
Startup Advisor, Keynote Speaker, and
Consultant to Fortune 500 merchants
Connect with Simon Taylor | LinkedIn
Founder FintechBrainfood
Market Development at Tempo
Advisor at Sardine
Connect with Tal Link | LinkedIn
Expert in Risk Management, Governance, and Controls
Connect with Hailey Windham, CFCS | LinkedIn
Host of the Fraud Forward Podcast
Banking Community Lead at Sardine
Certified Financial Crimes Specialist (CFCS)
2023 Credit Union Rockstar, CU Magazine
Continuous Improvement Award, SAFE Federal Credit Union, 2023
Top 20 Professionals Under 40, The Sumter Item, 2022
Guests
Episode transcript
Hailey Windham
00:05
Tal, thank you so much for agreeing to come on to Fraud Forward, uh, here live at Sardine Con.
Tal Link
00:11
Thank you so much for having me. I'm very happy to be here.
Hailey Windham
00:13
I am equally as happy to have you on.
Hailey Windham
00:15
Joseph, thank you so much for joining me on Fraud Forward.
Joseph Cox
00:18
Yeah, absolutely.
Hailey Windham
00:19
My favorite, favorite uh podcast host ever, joining me again. But this time we're live in person. Which is always my favorite thing. So, we are at Sardine Con in San Francisco. Karisse, thank you. And welcome back to Fraud Forward.
Karisse Hendrick
00:32
Thanks, Hailey. I am happy to be here.
Hailey Windham
00:35
Simon Taylor, let me tell you, uh, this moment right now is bucket list check item. Thank you so much for joining me.
Simon Taylor
00:41
Likewise. Thank you for joining me. And thank you for being part of the team Sardine, Sardine Media. I think you're crushing it right now. You're doing a great job.
Hailey Windham
00:47
Thank you so much. Means a lot coming from you. Truly. Um, I will tell you, I was kind of fangirling. You walked by this morning and I was like, "Simon, hi, my name's Hailey." And you're like, "Yeah, I know." And so,
Simon Taylor
00:57
You're famous.
Hailey Windham
01:00
No way. Not nearly as famous as you are. And I just really appreciate you taking the time to be here. And obviously crushing it with MCing this whole event. Uh, loved your opening. Um, and I'd love to just go ahead and dive in if you're good with that.
Simon Taylor
01:11
Let's go.
Hailey Windham
01:14
So, we are moving incredibly fast with AI. I think that's kind of an understatement really. Do you think we actually understand what is happening yet? Or are we still just too close to it?
Simon Taylor
01:24
No. I think for most people AI is that annoying chatbot, um, that won't just rewrite the email. And that you're shouting at. And I think for like 0.01% of people it's magic. And they're doing incredible things with it. And the big question is: yeah, but how does that actually help me manage this new fraud pattern? How does that help me with this new AML risk that's popped up and like this crime ring? And I don't think we've connected that fully yet. I think we're getting close to that, but that's the gap.
Hailey Windham
01:53
Yeah. So true. Um, and it brings up another point that even I've thought about just with like banking and fraud fighters with like that knowledge gap. Um, but this generation, right, is going to grow up with AI as something that's completely normal to them. Uh, we grew up learning how to search for information while they're growing up being able to simply ask for the answer. Potentially removing some of the struggle that taught us how to think and solve problems. Uh, does that ultimately put this generation at an advantage or disadvantage?
Simon Taylor
02:20
I think a disadvantage. Um, ultimately. I mean there's been a number, numerous examples of uh professors doing tests. Where the kids weren't allowed to use AI. And their grades just all dropped sharply. Uh and this is a consistent pattern. That like actually your real grades are 30% behind where your grades would have been 10 years ago, uh if AI had never been in the place. And you know this generation's been through the pandemic and now this. It is hard. Uh and I think that that it's sort of, almost creating this uh additional premium on people who do have industry expertise. Because that's becoming so rare and so hard to find. Uh but this will cost, correct. I for the life of me cannot do arithmetic, but I can use a calculator. And I'm functioning in society just fine.
Hailey Windham
03:10
Yeah. So true. Uh what skills become more valuable in an AI first world? Is it still about what you know? Or is judgment, curiosity, and knowing what to ask becoming more important?
Simon Taylor
03:21
Do you know I've had it a few times this week where I couldn't remember the name of that person, but my AI agent could. Um, but I knew I wanted to connect somebody like that, but I couldn't think of the name. So, uh, I don't think it's necessarily what you know, but it's the patterns you see in the things that you know. Uh your your brain is a very interesting beast. And everybody's works a little bit differently. And so you steer into what makes you, you. Uh, and steer away from, you know, just kind of turning the turning the handle every day.
Hailey Windham
03:49
Yeah. I will say I recently used it to help me, uh, decide what car I want to purchase. Because I was really struggling. I was like I don't know. But I needed to have all of these things. And no Google search, or car dealer, would help me. Because, you know, they only want to promote the cars in their shop. Not, you know, somewhere else. So,
Simon Taylor
04:06
This is why I'm secretly hopeful ChatGPT never does advertising. Because it's it's really good at getting you, like a well-researched answer. When you know you need the thing, but you don't know which of the things you need.
Hailey Windham
04:18
So true. So true. Final question. You challenged everyone at SardineCon to find something today that they can use tomorrow. What's one thing you've learned through your experience about AI that you think people should start using tomorrow?
Simon Taylor
04:30
Tomorrow. Start having a good old ramble at your AI. Click the little microphone button and say, "I'm kind of thinking of writing this long email, but I'm worried that it's going to come out like this, and I really need to say these things, but I want it to be kind of short, and I am also thinking it's Tuesday, and maybe we should send it in the afternoon." Just say stupid stuff at it. Say everything that's in your mind, and then say other things, and say more things. And say things that you didn't expect to need to say. Uh, and you'll be surprised how much better the output gets.
Hailey Windham
05:01
I love that. Thank you so much for joining me on Fraud Forward. And of course for joining us at SardineCon.
Simon Taylor
05:06
Thank you guys.
Hailey Windham
05:06
Would love for you to just give a quick intro of, you know, who you are, what it is that you do. So since this is the first time, but definitely not the last time that you'll be on the podcast.
Tal Link
05:15
Absolutely. So I'm a payment risk expert. I've been in the space about 17 years. I've seen everything from payment fraud to chargeback abuse, friendly fraud, credit risk, um fintech issues, and I really enjoy consulting on the space. I love hearing about people's problems. I love solving those problems. And, um, I think the space is just so fascinating and it's been a lot of fun building a career in this space.
Hailey Windham
05:42
Sounds like you're the risk therapist, you know.
Tal Link
05:45
A little bit but not so much. I mean, we have awesome companies such as Sardine helping with the actual heavy lifting of problems.
Hailey Windham
05:52
So true. So true. So obviously we are here at SardineCon. I'd love it if you could just uh, give us your first initial thoughts of it. Um, I believe we mentioned before this is mine and you're both first years, uh, attending SardineCon. So what were your thoughts?
Tal Link
06:06
Yeah, so first year. I am so impressed by the caliber of the talent. All the speakers that have come on to the stage today have spoken with really helpful and interesting tidbits. And story, and and their storytelling is really good. Um, I feel like I'm walking away with a lot of really practical solutions. And really good advice that I can now implement into my own work.
Hailey Windham
06:31
I love that so much. I think that the practicality of insights that can come back with you. You know, we we show up as practitioners and I'm now speaking, you know, as a former practitioner. But showing up at a conference and being able to take something tangible home with you that's like, okay, in theory, yes. Other things would work. But for me, what I can do today is this. And being able to truly just, again bring something back with you. Instead of just leaving with I mean, yeah, there's a good conference. There's a lot of interesting information. But something that you can really use, is so important whenever you go to these things.
Tal Link
07:09
I feel like I'm able to take a lot of the takeaways. And, um, apply them in a way that I can update rules and workflows. And, um, figure out how to prompt questions better, so that I get better AI results. And so that's been really rewarding.
Hailey Windham
07:26
Yeah. Was there a particular session that was maybe your favorite from today?
Tal Link
07:30
I'd say Soup’s session was my favorite. It was so detailed. And so relevant to everything going on. And I really agree with his narrative, which is we have to fight agentic AI with systems. Um, the fraudsters have access to AI. We have access to AI. We can't be doing manual reviews. We can't be scaling our human risk operations team. We need systems. Uh, workflows. And, um, to leverage all of the capabilities of AI to be able to effectively respond to all the fraud attacks that we are now seeing. I don't know if we're ready for agentic AI fraud.
Hailey Windham
08:12
There was a session that it is everywhere.
Tal Link
08:14
So, uh I, uh that keeps me up at night. So, I think Sardine is doing the right things to help prepare us to be ready.
Hailey Windham
08:23
And I think one of the things too that's highlighted, you know, throughout even today, was that there are systems that that we can put in place. But they're not meant to replace our our analysts. They're meant to help us be more efficient. They're meant to allow us to do more than what we ever thought possible. You know, right now whenever we're building a fraud program, if we are sitting back, it's like, ah, this seems like it's an unrealistic expectation of what we can prevent, what we can do. But if we think with our same team, with having this this big system or this solution behind us that could support us. Um, is you know I think the ultimate you euphoria where we need to get to.
Tal Link
09:03
I think that's exactly it. I don't think rules are going away. I don't think humans who are writing the rules are going away. But I think there's a way to turn these system, to leverage these systems, to bring human insights. And rules, and models, and AI. So that all of it together creates really great output.
Hailey Windham
09:22
So true. So true. So there was another part during SardineCon where we could all choose our own path. Um, I originally was supposed to be at a round table. But then ended up going to a fireside chat, which was great. Um, and I know that you went to the other fireside chat, which was about building a foundational model for risk. Was there any good takeaways that came from that session?
Tal Link
09:42
Yeah, one of my favorite things that I learned there, was that building models takes a lot of computing power. It's something that cannot be done in fractions of a second, or even minutes or even hours. It takes a lot of data, and a lot of effort, a lot of diligence. Um, data science is actually a science. And so one of the learnings I had was that it's very it's, a practical learning. Um, to run a lot of the models on smaller data sets. Confirm that whatever it is you're trying to do will work on a smaller data set. And then extrapolate that to a larger data set. So that helps reduce computing efforts and helps, uh, speed up the process.
Hailey Windham
10:26
I love that. Is something that truly is like, we we struggle because if we're looking for this one thing, but we've got too much data that it's combing through. It's like, number one it's going to take too long. It's not going to give you the same output. That and, it's not going to give you the opportunity to tweak it as well. But having that smaller data set, you're exactly right. Will will allow you to enable it better. And then be able to put it at scale.
Tal Link
10:48
Yeah. And I think AB testing too, is also very important component of that as well. So they talked about that a little bit. And and not just that session, it was talked about throughout the day. So I think AB testing in general is something that the industry moved away from a little bit. Don't know why. It's so powerful. And so, uh, kind of a gentle reminder to bring it back was also really nice.
Hailey Windham
11:08
So true. Love that. Um, any other thoughts about SardineCon or you know just the day in general?
Tal Link
11:15
So yeah, one of my favorite things was having the opportunity to go see the demos. I saw a demo about Sardine's workflow feature, or workflow product. I don't know how you guys brand it. But I was impressed because I liked that you could turn a workflow into several workflows, or ingrain other workflows within it. And you can really build it to be nuanced, and custom, to whatever you need to do. So it could be at a different life cycle, it could be for a different geography, it could be for a different, uh, product or feature. And so it came across how powerful of a tool that really is. And I was not expecting it to be as powerful as it seems like it is.
Hailey Windham
11:56
Uh, I'm continuously amazed by by the team here at Sardine. And and what we've built, and continue to build. You know, we've got a a weekly meeting with our team. And it's always something new that they've discovered or have been able to implement. And it's uh smartest people I've ever met.
Tal Link
12:12
Seems like it for sure. Hailey Windham Uh well, thank you again so much for joining me on Fraud Forward. This is the first but definitely not the last, right, that we'll have you on.
Tal Link
12:20
I would love to be on again.
Hailey Windham
12:22
Awesome. Great. Well, thank you again so much. And uh we'll see you and see you next time.
Tal Link
12:26
Thank you. See you next time. [Ad Break (12:26): Finally. I'm so happy to share with you all that The Saturday Fraud Strategist is now a podcast. What? Yeah. On top of my weekly newsletter, you could now listen to and watch me talk about my and hopefully your favorite topic, fraud strategy. And from time to time, I'll be hosting operators and founders to discuss where the industry is headed. And what we, fraud fighters, should pay attention to. I must say, I'm super excited and if I'm being honest, a bit nervous about all of this. I've been debating with myself whether to start a podcast for ages, but kept putting it off. But now this teaser is out. So I guess there's no turning back. So if you want to join me for the ride, head over to Sardine's website and subscribe now. Are you ready? Am I ready? We'll find out, next Saturday.]
Hailey Windham
13:32
So your session, uh, is, was titled we're not ready for AI powered fraud. I obviously completely agree with that. Um, but that is a pretty definitive statement. What have you seen in your reporting that convinced you that we're not ready?
Joseph Cox
13:46
It's mostly because it wasn't just discovering this, these technologies like the fake ID generation or the real-time deep fakes or anything like that. It's that I actually went and sort of deployed them against real targets. So making a photo of a fake ID, then going to a cryptocurrency exchange, and successfully verifying quote unquote verifying my ID with this fake documents. Like oh well we're clearly not ready. Because I, I, because I did it. Now of course maybe changes have been made, and I can't do that anymore. But like it's going to be a constant thing.
Hailey Windham
14:18
Yeah. Uh I totally agree. And, you know, with you taking that next step. Where it wasn't just, uh you know, generating an image. But you actually did do the testing, um, to see what the tools could do up against the real, like quote unquote verification process. Um, what did you learn from taking that next step?
Joseph Cox
14:37
Uh that it's practical. And it's real. You know, like it's one thing to see this criminal technology and be like, "Oh, that's pretty cool.” Or neat, or novel, or terrifying, or whatever. But it really doesn't mean much unless you can actually deploy it. And sort of take that next step. There's all sorts of like, theoretical attacks that cyber criminals could do. But they're kind of meaningless, if we don't actually see if they actually work,
Hailey Windham
15:02
Right? I'm I'm always one of those hands-on learners, too. Uh, where if I don't see it uh in action, it feels like it's theoretical. Um, but whenever we are able to see the deep fakes in action, whenever it's happening in real time, it's like, are our processes that we thought were supposed to help prevent fraud, are they even worth it anymore? Do they work at all?
Joseph Cox
15:22
Right. Right. I mean, there's going to be that constant cat-mouse game. Where people are clearly working on this. And like, obviously I don't I don't work in the financial industry, or the cryptocurrency industry. There are a lot smarter people actually tackling this problem. I know they are trying to tackle it. But it's going to be just this constant thing of, oh, XYZ company introduced this safeguard. Well, now the Southeast Asian crime gangs have discovered this tool or whatever. Like, it's never going to stop basically.
Hailey Windham
15:50
Plus, they don't have the regulators and rules to follow like we do.
Joseph Cox
15:53
Yeah. I don't think Haotian AI, which is the name of the tool I use, I don't think they really care about, uh, legislation and regulation. No.
Hailey Windham
16:01
No. Yeah. Um, what uh scares you more? Is it how capable these tools already, are or how quickly they're improving?
Joseph Cox
16:07
It's actually more how accessible they're becoming. Because whereas before people have been making deep fakes, and to a certain extent real-time deep fakes for a long long time,. But they've needed a very powerful PC, they've needed the technical knowledge to do that. Now you can be a non-technical criminal and just pay somebody for a service. Basically like, in the same sort of way that a company just buys a SaaS product, right? Criminals are doing that as well. And that's a real real advantage to these criminal groups. They don't have to sort of reinvent the wheel every time. And that's what concerns me most.
Hailey Windham
16:41
Yeah. Um what I don't want from this particular, uh, part of the episode, uh, to do to fraud fighters listening, is to have them walking away thinking, "Well, we're screwed." But. Um, where are you seeing defenses that can actually keep pace?
Joseph Cox
16:56
Uh, I mean broadly, I do think some companies are able to detect that, oh that is a deep fake. Like it may look very realistic, but I do think broadly AI detection is getting somewhat better. And there's a parallel with you know a AI generated text, and that sort of thing. And then AI companies also watermarking their own technology, and their own output, and that sort of thing. I do think some companies are getting better at identifying deep fakes. That's good. That's going to stop sort of KYC bypasses, that sort of thing. That's not going to help your family member, or even yourself, on a real-time video call necessarily. Because that's up to you as a human to decide. And that like I don't know how you fix that.
Hailey Windham
17:38
Yeah. How do we, um, give the gut to AI? Right. I always say you trust your your spidey senses, or just that gut feeling. Um from, this just doesn't feel right. Every, it checks the boxes. The verification was there. Somebody did click a button, but still just doesn't feel right. How can we get the gut in there?
Joseph Cox
17:54
Yeah. I think people just need to listen to that. Still, even while everything else becomes way more convincing. The voice stuff, the video stuff, the ID stuff. We can't then just get complacent and go, "Oh, well it's real now." You know, no, we actually have to be more vigilant now, because everything could be uh fake. Yeah.
Hailey Windham
18:13
Right. So true. You know, Simon Taylor opened SardineCon by asking everyone to find something they could use tomorrow. So, I'll give you the same challenge. If I'm running a fraud program and listening to this tomorrow morning, what's one thing you want me to do differently?
Joseph Cox
18:26
It is the detection of the deep fakes, but not just on the KYC side. Like it, should, there should be tools. For individual people and users, to be able to use, where they can go. They they log onto a video call, or whatever it might be, uh or a verification system, and they're the ones being told this is fake. It it's one thing for companies to protect themselves,
Hailey Windham
18:47
Right?
Joseph Cox
18:48
But then there's another thing to protect users and like ordinary people.
Hailey Windham
18:51
So true. Well, thank you so much for agreeing to hop on the podcast with me. I've really appreciated your insight and enjoyed your session thoroughly.
Joseph Cox
18:58
Yeah, absolutely. Thank you.
Hailey Windham
19:00
So obviously SardineCon has just, uh, I I can't even describe it. The the vibe today has just been amazing. I am completely on brand with my all purple self.
Karisse Hendrick
19:10
You are.
Hailey Windham
19:11
Um, and just the atmosphere here. Everyone is talking. I'm hearing collaboration. I'm hearing innovation. I'm hearing, uh you know, collective innovation. Which is really cool, too. Where it's, you know, that combined effort of, we can't just stop and we can't keep things in silos. We got to keep moving on. I'd love to get your take on, you know, Sardine Con 2026.
Karisse Hendrick
19:33
Yeah. So, I attended SardineCon 2025, and it was a lot smaller. Um, probably half the size. I mean, I know, that they had like several hundred people come. Um, this time around, maybe around 300 or so. So, um, much bigger space. I love the venue. I literally can see the Golden Gate Bridge from the window right behind the camera. Um, maybe I'm just going to stare at it for a second. No,
Hailey Windham
20:01
I keep getting distracted by it.
Karisse Hendrick
20:03
I know. Well, it looks like a mirage because it's behind fog. Um, but it's beautiful. Uh, yeah, I saw a lot of familiar faces, but a couple of new ones. Um, I think the the theme, so last year's theme of SardineCon was The Scamdemic.
Hailey Windham
20:21
Yeah.
Karisse Hendrick
20:22
And Erin West spoke, and like the whole theme of the conference was around scams. And that felt right. Uh, and this year it's around, you know agentic, not just commerce, but just like AI everything. And I think that's also on brand for this year. Um, even six months ago we weren't as far with AI as we are now. Um, and so are the fraudsters. They're also, you know, much more advanced than they were 6 months ago. So, um, I really enjoyed the Joseph Cox, the founder of 404 Media. I subscribed to their newsletter. I really enjoyed their um, investigative journalism around cyber crime and fraud. So, um, I enjoyed his talk. Even though a lot of it I knew about already. It was still interesting to hear his perspective, and how he's speaking with cyber criminals, and getting how he gets his information. Since he's not really in fraud. There's kind of the two sides of AI, right? There's the agentic risk operations side and then there's the side where the fraudsters are using AI to attack us. So, that's really been the, the two sides of that have been the biggest talking points, I think, uh, at the con.
Hailey Windham
21:44
Yeah. So, true. I enjoyed both of those. Actually, uh, Joseph was on the podcast earlier. So, yeah. So, we'll be able to to see some of that. Got to um, kind of deep dive a little bit more where I was like, there's just a couple of questions that I have. Uh, you made me curious. Uh, which, you know, we're we're in this every day. We see these things and we are talking with, you know, practitioners that are in it. That are literally in the fight every day. And so we kind of have that familiarity, but then to get just another perspective, especially for me. You know, I talk to a lot of community bankers and credit union representatives, but being able to talk to someone who's on that side, really helps give at least another uh, you know, bird's eye view of what it looks like. Because again, I, you're one of the main people that taught me this, but having that holistic understanding of fraud. Where it intersects, where we could potentially collaborate together. Really shows, you know, how how far we can go, and and what we can do in order to prevent, mitigate, detect, all those great things.
Karisse Hendrick
22:45
Yeah. It's great to get the on the ground perspective. But then it's also really good to get that 10,000 foot view of a solution provider, of a vendor, of a journalist. Uh, who's seeing all of it, and how the pieces connect with each other.
Hailey Windham
23:01
So true. So true. Um, and so I definitely want to ask about your round table, uh, session. So there was at one point of the, uh, conference where we were able to, uh, break away. We can kind of could choose your own path. Um, there was an opportunity where I was supposed to be at a round table but eventually, I mean I ended up, um, on the main stage. Uh, so, and but it turned into a podcast episode. So that worked out great. You guys will get to see that. Everything is content. Um, but your, uh, specific round table was on agentic commerce. I'd love to know, you know, what kind of conversations were had around that table for those 45 minutes.
Karisse Hendrick
23:38
I was really surprised that it was so full. There were two large, uh, round conference tables put together. And they were full. There's so many unknowns about agentic commerce right now, that a lot of people are just speculating. And so it was kind of hard to understand, to know in advance what people wanted to talk about. So I asked everyone, you know, to share, to introduce themselves, and to share, you know, their name, their company. And then what about agentic commerce they were hoping to learn. Um, and for the most part, I would say all but maybe one were all on the same page. You know, where they're just like we we want to know, like, what we don't know. And it's like, well I don't know if any of us have the answers. Um, there's so many, I don't want to use the word war. But there's so many, um uh, conflict, conflicting information or whatever, around protocols. The card brands want to own the protocols. The AI companies want to own the protocols. In, on the consumer side, you know, who's going to own the liability of the disputes? Especially, who’s going to ensure that the thing that was purchased was the right thing?f Just a lot of unknowns, so we talked through those. It was really fascinating. We had someone from one of the two card brands, uh, at our table. And he shared, um, some more specifics around some of what the protocols can do. Um, early stage, the ones that are out now, versus the ones that they're working on in the future. Um, that was interesting. And then, uh, I think what we kind of summed up at the end after like talking with everyone, someone said okay I think what I learned was B2C is nowhere close. Like agentic commerce for B2C we're not there yet. But B2B is possible. And there was one guy, uh, that represented a startup that facilitates B2B agentic commerce. And it's without the money, so the money gets transaction, transacted at end of the month. But that they have agents talking to each other about, um it's like real time invoicing almost.
Hailey Windham
26:08
Okay.
Karisse Hendrick
26:08
Um so there's not an actual transaction, but they're you know pinging each other for like, real, in real time. Um, so that's a good step in the right direction. Um, but yeah, it was it was different. Everyone was coming at it from a different perspective.bSome people had payments expertise, some didn't. Um, yeah.
Hailey Windham
26:30
Yeah. Did you ever or I kind of feel like I know the answer to that, but did it feel like the conversation kind of went too quickly? And then like, you were out of time.
Karisse Hendrick
26:39
Yeah, for sure. At the start of it, I was like, how am I going to fill 45 minutes? Um, and then at the end I was like, oh, I think everyone else is getting up. I think we need to go. Um, so yeah, I mean I posed a couple of questions based on guests I've had on the podcast recently. Especially um, I had Dave G on the podcast. Who, uh, worked for Google for 12 years. He's now in the crypto space working for one of the largest crypto companies. We're not allowed publicly to say which one. You can find it on LinkedIn. Um, but uh, he had said on my podcast that he didn't think agentic commerce would be about shopping for shoes, or shopping for a trip, or shopping for a concert, or whatever, those big purchases. He believed that it would be for milk, bread and eggs. Things that you just want to show up automatically. Um, or he also floated this idea of micro payments on stable coins. And so we talked a little bit about that. And the Mastercard guy shot down stable coin idea real fast, which was fascinating. Of course he would. Um,
Hailey Windham
27:52
Of course. Why is that?
Hailey Windham
27:53
Right. Right. But it was in, his explanation was interesting. And you know, uh, but then the whole, you know I asked about, because Dave's whole point was humans aren't going to want to get rid of the retail therapy part. Part of online shopping is the hunt. The hunt for the deal, the hunt for the collectible, the hunt for the best trip. Putting together each part of your trip. Some people really like that. I don't, but some people do. Um, and
Hailey Windham
28:25
Find me an antique though,
Karisse Hendrick
28:26
Right? Exactly. Yeah. If you're looking for like a specific
Hailey Windham
28:30
Hailey puts a hurt on that one.
Karisse Hendrick
28:32
I know. And I think I was the one that told you.
Hailey Windham
28:34
Yeah, you did. Uhhuh.
Karisse Hendrick
28:36
My husband appreciates. Yeah, wait till he hears about my next, uh my next addiction. Um, but yeah. So that, he was posing that humans aren't going to want to have an agent buy those big things. But that they will want an agent to just buy milk, bread, and eggs. Like the boring stuff that they just need at their house when they're out.
Hailey Windham
29:02
Yeah.
Karisse Hendrick
29:02
Um, so I asked that question to the group. Um, there were differing answers. Um, I think that the card brands are really thinking about it in terms of large transactions. Um, but they haven't really figured out the dispute part yet. Uh, they haven't figured out the acceptance, the authorization piece of authorizing a credit card through an agent. They also have not figured out the dispute part yet. So, those are the two pieces that I think for B2C, we're just not there yet.
Karisse Hendrick
29:40
Yeah. I feel like, um, and I I keep thinking this. And I know it's not an easy fix, but like, we've got the MCC code. We've got the E ECI contactless code, that lets us know how a card was, you know, used. Why can't we have that for those agent purchases? You know, like it, let it just be one other data point. But I know it's not there yet. And you know, hopefully we can get there. Because that was always my mindset, whenever I'm thinking practitioner role. Alright, how would I prevent agentic from happening? We're not ready for it. And the only thing I can think is, what's the code that I need to block? You know, what data, what data column do I need to know means that an agent did this versus a human? And we just don't have it yet.
Karisse Hendrick
30:23
Right. Uh, there are some very large merchants that have done several tests with all the main AI components and and tools, uh, to make purchases on their own website. And then they go back and look at all the metadata. And currently right now there is no metadata that advises them that it's an agent, when they're using their traditional tools. I know that Sardine, for example, does have a way to identify if an agent made a purchase. But, um, using what they have right now, they they have looked at every single data column, and there's not a single one. I think Visa and Mastercard should one of their or both of their protocols win out this war. Um, I do think that that will be in the descriptor, that will be in the payment, uh, you know transaction string of information of data, that you know gets picked up through the authorization cycle. But, uh, that requires more integration for merchants to be able to accept that new, you know, descriptor code. And just, and and be able to put it into all their data and their reports and everything. So, we, I still think we're a little ways away. But I know that they're all working really fast on those things. Um, because there's a lot of money there.
Hailey Windham
31:48
Right. Right. Well, Karisse, thank you again for joining me on Fraud Forward. Was there anything else from SardineCon that anyone who wasn't in attendance should know about?
Karisse Hendrick
31:59
Is this where I call out Matt Vega for
Hailey Windham
32:01
Yes, please.
Karisse Hendrick
32:03
For standing us up.
Hailey Windham
32:04
I I got on stage and I said, um, I can see some some some confused looks on your faces. I'm the prettier version of Matt Vega.
Karisse Hendrick
32:13
I’m like, is this where you want me to call out Matt, for calling out like an hour before? Um, no. I think it was a really good time. There's, the, it's a very unique group of people. I I kind of wish it was a two-day event so you could meet more of them. Um
Hailey Windham
32:31
2027,
Karisse Hendrick
32:32
Right? Yeah.
Hailey Windham
32:33
We'll go ahead and place that bug now.
Karisse Hendrick
32:36
Eduardo is going to kill me for saying that, because one day is enough. But it is a lot of planning. But, um yeah. I just, um, I think I've had some really unique conversations with people. I just was talking with a guy that, uh, I know a lot of the people he used to work with. And some of the projects he used to work on at a very large retailer. Um, so we were able to talk about those things. Um yeah, it's just been a good time. And you know, great to see familiar faces, great to meet new ones, all that.
Hailey Windham
33:05
Absolutely. Well, thanks again Karisse.
Karisse Hendrick
33:07
Thank you Hailey.
Hailey Windham
33:14
Thanks for listening to Fraud Forward. Remember, every conversation, every connection, and every insight moves our industry one step closer to stronger fraud defenses. If today's episode sparked an idea, share it with your team or tag me on LinkedIn. I love hearing how you're moving fraud forward in your own organization. Until next time, stay curious, stay resilient, keep moving fraud forward.
View full transcript
