John Ridd is a technologist with a long track-record of taking both disruptive and transformative technology into unsuspecting markets. Currently he is CEO of eviid.
Are you ready for the unseen risk? Hence the missile is coming at you. So here we are. We’ve excepted that we’ve now got media. And we’re putting it to you use into every aspect of our lives. What we have done in 2017 we took 1.2 trillion digital images. That’s a huge number.
What we have done, is we’ve also started to look at how we use it in our everyday lives and excepted that as part of a process. So if we just reflect back on the fact that we’ve got this got all these images all around us in everything we do let’s first look and reflect on where we use it.
Well historically we used it by taking an analog photograph. And when we took a photography we had a negative, we developed a negative, and alongside that negative we printed a photograph. And we had the 2 opposing pieces where we could look at and check that there’s a reference point to verify that that was true.
Now it was possible to make some chemical changes. And if you knew how you could also take it to a specialist that could look under a microscope to see what that change was. But with digital images it’s different. But we’ve put digital imaging into everything we do. If you think about that first point is we’ve put it into drones dash cams GoPros.
We’ve put it into the home, into smart devices in the home such as Tvs. Fridges, freezers. Into the door bell. But it’s much deeper rooted than that. Because we’ve also put it into manufacturing. We’ve put it into food production. We’ve put it into healthcare. And we’ve put it into CCTV. And CCTV is capturing us an average of 70 times 70 times per day. That’s an astounding amount.
But let’s reflect on that in insurance. We’ve really embraced that into the insurance process. We’ve all been really quick to look at strategically how do we transform our business models? How do we create new revenue streams? How do we actually get faster traction in the market? At a tactical level. In a claims process. We’re then looking at it in terms of speeding up the claims process itself.
Saving time, saving money. Enhancing the user experience. ‘Cause ultimately, that’s what it’s about.
If we can improve that user experience it’s much more sticky for that customer to stay with us as an insurer. So that’s really important. So all these things are really important to about how we embrace the technology. And how we use it. But unfortunately it’s not that simple. You see whilst I was preparing this presentation somebody from FRISS asked me a really good question.
As we were going through some exchange of information, I said obviously there is a difference between certified and circumstantial evidence. And they asked me: Well, what is the difference? And that is the cornerstone of what I would like you to take away today from my presentation. Is this point:
Certified information…fingerprint, is unique to me. My DNA is unique to me. If we think about CCTV, this is great, there is different models you can think of in the technology sector, but CCTV is a great thing. It stands for Closed Circuit TV. So I can authenticate the image. I can authenticate with that camera and I can verify it through a process.
And it uses an encryption in there. A single-direction encryption called SH8265. So that means it can be verified. That it’s one consistent piece of information. It’s in a closed circuit TV. So what is circumstantial? Well, circumstantial is when you can’t authenticate something. So with an image, with a digital image can we authenticate the source? Does that source when you take that piece of information, can you authenticate it?
Now I forgot my little mobile phone here. But everybody’s got one. I was gonna pull it out. Everybody has a mobile phone. And everybody has a digital camera on it. Can you authenticate that that is true and accurate information?
So let me pose you a couple of questions here. Do you all know you can change the metadata on your photographs and your videos? Do you all know that you can Photoshop something? Do you all know that what you can do is you can click a video and change a video?
And you’ve all come across the fact that in the insurance sector that our concern is, you might have somebody that searches the internet to take a photograph of a washing machine leaking out over the kitchen floor and then somebody can take that and send you that as their claim. And we’ve got some barriers to prevent that.
But in the insurance sector we’ve also embraced apps. To help us improve the process of managing the claim. Or onboarding the claim. But in that app, have we got a way of certifying the media? And the answer probably is you’ve not.
So where we are, is we are at a point where we’ve introduced mobile phone based technology. But it’s further deep rooted than that we’ve actually introduced digital imaging into multiple aspects of our lives and we’ve no way of certifying it.
Because there is a human race in society. We’re fantastic in embracing technology and running with it really fast and then finding actually we’ve got holes in it. We all have seen what’s happened with Facebook. This is an important point for us. So this is an important point for me to hopefully get the message over.
What’s the difference between certified versus circumstantial? So you’re on this road with me so far, we’re at a point where we got one point on the axis where we can determine that with human intervention we can change the media. And we work with a law company. And quite interestingly they’re a large practice. And they said to us, you know, in 20 years we’ve been submitting digital images to court and we only ever had 3 occasions where the media has been thrown out of court. 3 occasions. But those 3 occasions are all in the last 18 months. You see defense lawyers have known for some time, and so have the courts that actually the media we’re looking at is just circumstantial information. It’s not certifiable information.
So on that axis we’ve got humans. And on the other axis the next point we’ve got in the crossover, that’s coming at us right now is the machine.
So this is a digital website you can go to. Hands up who’s seen this person does not exist. Ah, that’s interesting. It’s not that many. Fantastic. Take a note of this and go and have a look at it. ‘Cause this is your first step into seeing an AI system. And it generates a unique person a unique, fake person every 9 seconds. It’s generating somebody that does not exist, and it looks like it’s real. It’s not!
So you’ve got this intersection where, again, you can see Barack Obama, you’ve probably seen that on the news where they’re all concerned that Barack Obama was making a speech. But it wasn’t. It was an AI system. And that’s called deep fake. So you’re at this crossroad where we don’t know what we’ve got.
Can we trust it?
So what’s the current state of technology? So a little bit of an introduction to EVIID of what we do. We created a system that allows us to certify and verify the creation of media. And in our platform we’ve really focused into the insurance sector. By working through video solutions that allow policy holders to send information in. Or loss adjusters to send information and sharing it, whether that’s uploading video, or live video, or capturing videos from drones.
And in some of these solutions that we’ve taken to market we’ve got for example where we’ve put the use of video into high volume, low value claims into content insurance.
Do you know what we’ve had? We’ve had 40% of people walk away. Because when they actually start to see, I’m gonna use video, I’m not gonna fake a few photographs, I’m actually gonna start to use a video to show the damage. 40% walk away. If that’s not fraud, I don’t know what is.
We’ve reduced property claims cycle by 68%. We’ve reduced motor claims cycle by 98%. With the use of video. But the core and underlying piece of what we’ve done, is the fact we have this ability to verify and validate the creation of media. And that’s a patented algorithm both in the US and Europe. So we can identify, if you’re using our platform, deep fake. That’s a big claim for us.
So we’re all about how do we make this a standard. How do we take it to market? We’re at this juncture where we got human and we’ve got machine. But around the corner we also need to think about, in the hack, is the hack coming from somebody set in the home or is the hack coming from organized crime.
I think we need to take stock of where we are. Now this…How many people have seen this as a news article? This came up as a news article in the UK through the BBC. And it was from a team of Israeli white hackers, who hacked into a medical scanner and they changed the cancerous tumors.
So we have no knowledge here neither the AI system, nor the medical professionals could identify whether those were bigger, smaller, or even real. When you start to get to understand what’s not real and we actually got 1.2 trillion images. This becomes a big problem. So imagine if I’m an organized crime network. I could create a network of patients where I could fake the tumor. It could be a false negative, I could make it bigger, smaller, or even make it disappear.
But you put your own imaginations into that. I can start to build a claim network. And it haven’t got to be that deep. But the size of the claims are massive.
So where is the elephant in the room? Whilst I was preparing for this, I thought I sort of need to reconnect in terms of perhaps help you reconnect where organized crime comes from. Now there’s a European retailer that suffered a hack by an organized crime network in Eastern Europe. So out of territory, hard to get to.
But these guys are sophisticated. They got plush offices. They got all the big monitors. They got all the know boards. And the hackers know about how to get into the retailer to extract their bank information. Or even who’s using their bank. Extract credit card information. Now the hack was about understanding how to get into the supply chain. To extract. So what they’re doing, extract expensive items and goods. So what they’re doing is looking for key vulnerabilities in organized crime.
And in this process what they did s they were able to lock into the supply chain for nonstandard goods. Because the retailer wants you to sell a 72 inch tv. Or other items. So let’s just go through that. They placed an order. The order then goes to manufacturing. From manufacturing it comes into the warehouse. From warehouse it moves through to dispatch. And from dispatch it moves to the van. Whilst out in transit they could see that the crime network could see that, because they’re watching the whole system the whole time.
What did they do? Cancel the order. So the order gets cancelled, it means it comes back around into goods inwards. And what happens at goods inwards? It’s nonstandard, so actually I can’t put it back into the warehouse, ’cause it’s a nonstandard good.
And at that point the organized crime network had somebody in the warehouse and had somebody on the vans. And they were moving goods out. 50 million of goods disappeared.
So my point in that, is this is where we are with media. We’re at that point where we’re bringing it in. The organized crime networks now know with AI systems, and human intervention. This is gonna be a tough job to prove.
What we’re looking at is real. Let’s just lift that lid a little bit further. Because everything we look at is at risk. We’re accepting that we can use digital media in the mortgage market. And if you look at the mortgage market and if you measured out the UK, France, Germany, the Benelux countries it’s about 650 billion euros a year. Now a smart crime organization somewhere else could look at that and say, do you know what?
You’re using that for the collection through state agents, through the buyers, through the sellers, ’cause we’re using digital images to come on. E-mail. Or come through apps. But we can’t verify it.
How do we know it’s true? And we don’t. But these guys get smart. What’s 1 percent? It’s massive. And these for me is the part of the journey, this is what’s at risk. So what’s our vision? Our vision is about the creation of certified media. Of verifying that media. And working in collaboration with FRISS.
With their fraud detection systems and AI systems, where we can really deliver a unique proposition.