Live Blog | FRAUDtalks 2019

Sep 24, 2019

Wednesday, September 25, 2019, FRISS' annual event #FRAUDtalks took place in Amsterdam. 150 insurance professionals from different European countries gathered together and listened to six appealing talks of twenty minutes each. The live blog below gives you an overview of the event in reverse order. This live blog is closed now, but and there's more to come: CU @ FRAUDtalks 2020!

Wrap-up FRAUDtalks 2019

16:50 hrs - Pep Rosenberg ends the FRAUDtalks conference with a humoristic wrap-up of the afternoon. Thanks to FRISS and their staff for organizing this great event. Let's have a drink in the Amsterdam Tower, and after that, travel home safely! Many thanks for coming to our #FRAUDtalks.

Viva #FRAUDtalks 2020!

The IT perspective to succesfully finish an insurance fraud prevention project

Bennett Arron - FRAUDtalk 6

16:35 hrs - Some time ago, writer and stand-up comedian Bennett Arron were in serious debt. He owed thousands of pounds to phone companies, banks and department stores. Only, it wasn’t him. Bennett was a victim of identity theft. This theft resulted in Bennett becoming penniless and homeless. Years later, he used his own experience to direct and present the TV documentary ‘How To Steal An Identity’. Due to the story told, Bennett was taken in custody. Instead of seen as an eye-opener to the public and the government, the documentary was seen as a means of distortion. Bennett is at #FRAUDtalks to tell what it’s like if someone steals your identity and how devastating the consequences are of making a documentary in the public interest that reveals an inconvenient truth.

Kafkaesque experiences

Bennett starts his talk with: I'm Jewish and I'm from Wales. I thought this was my identity. But it appeared that my identity was something totally different. Twenty years ago, when my identity was stolen from me, I found myself back in a Kafkaesque process fighting against companies, banks, mobile service providers who accepted a customer that wasn't the man he said he was, namely me, and I wasn't it. While making the documentary I asked for people's personal information. It was unbelievable how many people gave all their passwords, pins, their mother's maiden name, you name it. After that, I made up a new identity as if I was the State Secretary. It turned out to be very successful, but I got in a lot of problems afterwards. These days I make my money by telling my story to guys like you. Thank you!

Jeffrey Rapattoni - FRAUDtalk 5

16:10 hrs - As an attorney, Jeffrey Rapattoni focuses his practice as co-chair of the Fraud/Special Investigation Practice Group of Marshall Dennehey on insurance fraud, bad faith and Special Investigation Unit’s (SIU) related matters. At #FRAUDtalks Jeffrey shares how insurance professional can use data. There are a number of ways depending on how you choose to use your data. When the full resource of a carrier to view a potential fraud event and confirm or deny the event in less time, the insurer will be able to spend more meaningful time on tasks that need attention.

Data in the insurance industry

In the insurance industry nowadays,  the most relevant question is:

What difference has data made in the insurance industry?

I turn the question upside down.

How do you intend to stay relevant without using data meaningfully?

Our industry in the US was the last of all industries to embrace the power of data. Currently, we have a New Now in which we do use data. We're moving forward. The problem is that claims automation enhances fraud. If data is not actively used in your process, fraud detection is too difficult. The problem is that it's not something that can be introduced easily, it takes time.

Round table

My best advice for you is to set up a Round Table to get to the New Now: bring in your IT department from the start. They need a seat at the table. Claims must have a seat because it's their business, and the have the money to pay for the investments. Finally, SIU must be at the table. Sometimes SIU and Claims fight each other, but the advice of SIU will be investigated by claims. They mustn't fight each other, they need each other. Data doesn't replace the human touch, data enhances the human touch. People will always be needed to judge fraudulent claims. When you collaborate in renewing your process, two factors are most important: collaboration and innovation. You must be aware the innovation is meaningful to your company, don't copycat, but think and rethink what will suit your company. Data is taking you to the New Now. Claims automation is inevitable, so bring your carriers up that future state.


Ynzo van Zanten - FRAUDtalk 4

15:30 hrs - Tony’s Chocolonely, who doesn’t know these famous Dutch chocolate bars? Their CEO aka Choco Evangelist Ynzo van Zanten will tell us more on Tony’s innovative roadmap to 100% slave-free chocolate and how anyone can contribute to fair trade business goals. Ynzo will address Tony’s Chocolonely, contemporary slavery, entrepreneurship, sustainability and leadership. The good news is you're going get home with chocolate, and the bad news is that you will go home with awareness.

The bitterness of cocoa

The bitter reality of chocolate is what I'm telling you about. At one of the sides of the supply chain you'll find the small cocoa farmers in Asia, in Latin America, but mostly in Western Africa. On the other side there is us, who eat chocolate without any feelings of guilt. In the middle part are some big multinational companies like Mars Inc, Ferrero Group, Mondel?z International, and Nestlé. Ynzo shows the audience some young African workers who worked without any payment and with horrible circumstances on cocoa plantages. To fight this kind of child labour the Harkin Engel Protocol was agreed upon. Unfortunately, the Harkin Engel Protocol was, and is a non-binding agreement. It didn't change the labour circumstances at cocoa plantages at all.

Tony's Chocolonely

Some years ago, Teun van der Keuken, a journalist from the Dutch television program Keuringsdienst van Waarde bought ten chocolate bars, ate a bite of each of them and then turned himself in at the police. He wanted to be prosecuted as a criminal, being guilty of eating chocolate and therefor being responsible. It appeared to be too difficult to get a conviction. Subsequently, he started his own chocolate factory, which should be 100% slave-free. He started Tony's Chocolonely. Nowadays, we're the mosquito in the cocoa industry, we keep the big companies alert and strive for them to change the supply chain. Tony's Chocolonely shows you can make honest chocolate, and still have a product that sells and earns its market. Our successful company proves it can be done, even in a saturated market like the cocoa industry!


We aim to inspire other companies to be like us. Not only at the front side, but also to copy our backside of manufacturing chocolate. We're crazy about chocolate, yet we're serious about people. We work with 5,500 farmers in Ghana and surrounding countries in Western Africa these days. We're proud that Ahold's Delicata chocolate is now slave free, especially since one of the big cocoa firms makes it. We have 18 different chocolate's, so we're also innovative, and we're commercial as hell. But the way we run our business is innovative, transparent and we make fun. For instance, we have a bonus for every child that is born and gets the name Tony. Why do we make our bars unequally divided? It is to show that assets are unequally divided in the world, and we show the map of Western Africa on our bars. I want to end with a quote of one of the biggest contemporary thinkers of Europe, and thank you for listening to my story!

Once we know and are aware, we are responsible for our action, and our inaction.

Jean Paul Sartre

Social Media

In the meantime...






14:20 hrs - everybody enjoys a break to either follow a workshop on car hacking or to have a drink and enhance their network. The hashtag for the event is still #FRAUDtalks or #FRAUDtalks2019.

Ken Munro - FRAUDtalk 3

14:00 hrs - Over 15 years, Ken Munro worked in the Infosec Business as a security entrepreneur and industry maverick. He is a known car hacker and destroyer of 'Internet of Things' (IoT). In a TEDtalk, he showed how insecure IoT products are and how easy it is to hack them. He's an ethical hacker. Today, Ken reveals some security vulnerabilities and shares his expertise on stage.

Children's toys

Ken introduces a doll that can talk with your child. He hacked the doll, and now she can swear like a docker. There are many more toys for children that are not secure enough to be trusted with your children. The saddest part is that the manufacturer of the doll wasn't interested in making the electronic device more secure. He shows the audience a child's smartwatch that can track where your child is. But there were many security issues with this device. You could easily modify it into a stalker's instrument. It's really creepy what this unsecured thing for children can do. He tells more about other smartwatches. To hack them is often very simple. An idea for a hack: the UK doesn't perform very well with the Eurovision Songfestival. Ken shows the audience how easy random votes can be generated. This fraud is very, very easy, and Ken knows for sure the Netherlands will not win again in 2020.

Car hacking

In cars, al lot electronic functions are vulnerable to hacking. Think of the GPS, your smart-lock, or even de-immobilise the whole car so it won't run anymore. It's not very difficult to hack these so-called secured devices. Another example is about CCTV. Ken hacks lots of CCTV systems. Next example: hacking housing alarms. To hack most of them is simple as well. Not all of them, we found two who couldn't be compromised. Guess what I have in my house? Smart doorbells: you can see who is at your door. But a criminal can use it to get into your wifi network and identify all of your passwords. Remember, it's just a doorbell! Ken's message to all attendants: ask for evidence of security if you buy things with electronic devices in it!!!


Gian Luigi Chiesa - FRAUDtalk 2

13:35 hrs - FRISS’s Gian Luigi Chiesa is one of our talented data scientists. Currently, Gian focuses on Artificial Intelligence (AI), where he explores cutting-edge approaches to detect insurance fraud, involving Deep Learning. At #FRAUDTalks Gian shares his vision on a future with AI. He will discuss how and when you should trust AI. He will answer the question: ‘What is interpretable AI? And is it enough for us to trust it?’.

Interpretable AI

Gian starts with a question for the audience: are you aware of AI that's already surrounding you? For instance, the scenario of autonomous cars is much closer than you think, but can we trust an autonomous vehicle? We already have a lot of contact with AI-powered technology. For instance, many of us daily interact with smart assistants such as Siri, or Alexa. We trust AI machines when we can observe that they systematically make the correct decisions, like weather forecasting, or self-driving cars. But what if we cannot directly observe the consequences of our decisions? Think about a bank that refuses a loan, or an insurance company not accepting a policy application: we will never know whether we made the correct decision. In these cases, we have to make the AI machine more transparent. Some models are easily interpretable. Gian explains a decision tree, it is a simple model. For complex decisions, you need deep neural networks. They are very powerful, but not easy to understand. What can we do in such cases?

A black box revealed

Several techniques have been developed lately, which allow us to make the decisions of a black box model more interpretable for humans: one of this is called LIME: Local Interpretable Model-agnostic Explanations. How does it work? Gian gives an example with a Wiener-Dog (tekkel), the shape of the body and the shortness of the paws, is the most important factor in identifying it is a Wiener-Dog. We see the importance because if we blur out the body from the picture, the black-box model cannot predict it as Wiener-Dog anymore. Gian concludes by telling us how trustable and interpretable AI can build a bridge between humans and machines, so humans can understand machines better. Researching and developing new model’s interpretation and explanation techniques can improve our understanding of intelligence and more, in general, our understanding of the world’s phenomena, which ultimately is the goal of science.

John Ridd - FRAUDtalk 1

13:20 hrs - Presenting John Ridd means we introduce a technologist with a long track record of taking both disruptive and transformative technology into unsuspecting markets. He worked with the Information Assurance Advisory Council and is an expert in cyber risks and insurance. John will peek into the future taking a quick look at the technology available to us today.

Adaptation of technology

The insurance industry adapted technology quickly. An important question here is: what is the difference between certified information and circumstantial information? Nowadays, in insurance, media are part of the process of both underwriting and the claim process. Certified information is for instance DNA, and fingerprints. The information is unique and points back to just one unique person. If you think of CCTV, it can be verified. Circumstantial information is changeable, like for instance a picture in which you can change the metadata or the picture itself with Photoshop. We cannot certify all information we use. The insurance industry processes a lot circumstantial information. You must be aware it can be fraudulent. Look at, every minute a new not-existing person is been made. Do you know deep fake? This is to let Brack Obama say things he didn't say at all. In our EVIID platform, see our website, we help to make certified information from circumstantial information using visual information. We can identify 'deep fake', and we can find out how the original picture looked like. Be aware that organized crime is already in cyberspace. They look for key vulnerabilities in a process and hack it, and use fake media. Be aware of your processes' vulnerabilities.


Welcome to FRAUDtalks 2019!

13:15 hrs - #FRAUDtalks is orange all over the place, yet usually, orange fraud is something weird from the White House, but this event is about fraud in the insurance business. We start with the CCO of FRISS, Marc Mulder. Marc starts by asking for applause for the people who organized the event.  [big applause follows] Marc's own experience is that the insurance industry looks boring from the outside, but once you're a part of it you discover how lively and interesting it is, and how insurance people look for connections. FRISS started the #FRAUDtalks four years ago, to contribute to getting more connected with each other. If a new idea is brought to mind, it's easy to react with a 'no' or 'yes, but how much more fun it would be if you start reacting with 'yes, and'. Today some new ideas will be presented. Keep an open mind! We hope to have a lot of fun this afternoon.

Pep Rosenfeld - #FRAUDtalks moderator

13:00 hrs - At today’s #FRAUDtalks Pep Rosenfeld is our moderator. He is a seasoned host who makes the event energetic and more fun (we hope). Pep introduces our six keynote speakers to the audience. Pep Rosenfeld is Boom Chicago’s creative director and is one of their performers. During #FRAUDtalks Pep will use comedy and improvisation to help to make every FRAUDtalk stick. If you’re on twitter, please use the event’s hashtag #FRAUDtalks.

Before lift-off

12:45 hrs - getting ready for lift-off. Please, use the hashtag #FRAUDtalks whenever you're sharing your experiences of the event! We've arranged a fabulous line up of keynote speakers, but firstly Pep Rosenfeld will welcome all 150 participants, which number excludes the people from FRISS itself.