Meet Timmi Rysgaard, Product Owner, Brand Protection

We sat down with Timmi Rysgaard, Product Owner (Brand Protection) at RightHub, and discussed the challenges and future of brand protection. He highlighted the increasing threats of counterfeit goods, sophisticated fraudsters, and the exploitation of online sales channels, especially on social media platforms.

TR: I see a lot of pressing challenges, including the proliferation of counterfeit goods, and I also see the field evolving greatly in the future. There's also something about the increased sophistication of fraudsters which creates a need to keep up with change regarding infringements, but also change in consumer behaviour. Over the last few years, there's been a lot of growth of online sales channels, especially social media, marketplaces and standalone websites – the pandemic really created a shift here. The entire world was kind of shut down for a long time, which essentially meant that people would stay at home and spend more time on the Internet, and buy more on the Internet, which created a playground for counterfeiters.

The latest huge thing that happened is that people now are online more, buying online, buying through apps, buying through social media, which is just causing a lot of ways for counterfeiters to exploit unknowing consumers. I would say that a challenge for BP (brand protection) specialists today is that through this growth, you as a specialist, need to have an overview of a way larger array of websites and often also across multiple jurisdictions. So, imagine sitting as an in-house legal team in Denmark you go by Danish law and by European law - but then all of a sudden you start to see something popping up in Malaysia, then Peru, and then Nigeria and now you need to understand what rules and laws apply in those jurisdictions.

I would also say that the fields are vaulting in the direction of more advanced technologies from the counterfeiters to the point where they are definitely getting more sophisticated. Sometimes you see something pop up and it's clearly a fake, you have no doubt in your mind that this is a replica or brand impersonator. But with some of the latest advancements, especially AI and machine learning, it's very easy to make pretty good fakes. You can go and 3D print any given object essentially. I thought a lot about this question actually, since part of the company is in Denmark and we do a lot of work in Scandinavian design, which is essentially a very minimalistic style. And since it's minimalistic, it is also very fit for a 3D printing mechanism to just go and replicate it. So, the thing that makes Danish design stand out is also the thing that makes it be very easy to replicate.

I would say that BP specialists also rely more and more on technology. As I said earlier, you have increasing amounts of websites and jurisdictions that you need to cover and keep up with, and you just need to have a technology-based approach to it. Now, everyone has heard about ChatGPT and large language models in general, and AI just became a buzz word – and they are not only useful, but also needed. You have to cover so much ground to protect the brand, and you can't do that by hand – that's impossible. And of course, also something like blockchain - where we haven't been involved a lot yet - is definitely something that you can start to use to verify authenticity.

TR: I want to spend a bit of time talking about AI because it's such a wide concept. It essentially comes down to; what type of AI do you like? What's your intent of use with the AI?

You can use it to create some text, but you can also use it to alter or generate some images for you, which essentially then can be used as your own design. You can feed it with a billion images of well-known designs and then ask it to create something that's similar and then you're on the border of creating a design infringement right there – so we sense AI powered counterfeiting as one of the emerging trends. And you see it on some of the websites that are popping up on a daily basis, because right now you can go and create a website which just a few clicks, you don't even need to be able to code that much. So, you can set up a website that looks professional and you can have people come in there and spend some money and the only thing they receive is a thank you letter (if they are lucky), and also a bad customer service experience because they are not getting anything at all.

Another trend is probably “social media”. Again, very much due to the pandemic people went to social media to buy stuff. And it is just so rapidly changing - you could go to a social media site and post a ‘story’ where you are selling or advertising something, and then that something is now in your description or in your bio leading to a website that is not indexed by Google yet, but you can still go there since you have the link from this particular story. And now you can go and buy something. This means that if somebody has enough followers, they can reach an insanely large audience, and then disappear within 24 hours again. So, to stay on top of that, you need to involve some sort of technology in this because then you might just miss out on this short story, and that story and so on.

But then again, as I said before, we also use AI and technology. So, it's sort of like an AI versus AI world both for creating counterfeits, but also for tracking them down and removing them. AI is something that we use heavily in our line of work for image recognition and for spotting trends among product pages using similar keywords or phrasing. They might have different selling names and brand names, but if you link them all together, there's breadcrumbs all over the place and you can pick them up and see that, ‘Oh, that's actually one guy going under 100 different names’, which is pretty helpful. And again, you wouldn't be able to do that by hand unless you have an insane mind.

One thing I want to mention here as well is that the number of brand impersonators - where a company is claiming to be another company - is growing. If you go to a social media page, there's one that has a verified mark and then some others that have a verified mark – but within their profile picture. So, you think they are verified, but that's just the picture that looks like it, and the name is the same and so on. They claim to be that company which can be very detrimental to a brand's reputation because they might pose as customer service and then ask for confidential information and so on. And now you're in trouble because you as a consumer have a very hard time distinguishing between this and the actual brand. The last year or so it's been also very easy to create deepfakes in terms of faces and also to recreate voices - so you can have a face of a CEO or someone else and you can also create fakes with the voice if you have enough material to do that. It's just crazy to think about how you might impersonate a brand and pretty much do everything that they do. As a consumer, it's very hard to recognise, so I would say that's definitely one of the emerging threats.

TR: That really depends on your program's strategy. The “reason” why you are trying to protect your brand is to see less problems, but it could also be that you just want to increase your brand's value. It could also be that you want to reduce the number of complaints your customer service personnel will be receiving. It really comes down to the brand’s brand protection KPIs.

I would say that if you were seeing counterfeits or brand impersonators online, and now you see a reduction in the number of new appearances of these online - it's a pretty clear indicator that you're doing something right.

We have very strange conversations sometime with clients where they are dissatisfied with us finding less and less infringement on them. And I'm like, ‘Well, isn’t that a good thing?’. I know that from a client perspective, they go ‘yeah, I spend X amount every month and now we're removing less than we used to’ - yes, that's because you are less ‘ill’ or ‘targeted’. Which is a good thing essentially.

A key indicator could also be that you are reducing your legal costs - if you spend X amount every year on legal battles because a few legal issues are arising out of these infringements, which you have to actually go to court for or settle in some kind of legal dispute, and you are now doing that less and less, then that could also be an indication of succes. Then, especially brands within retail, they could see the success from a brand protection program being that they are increasing their revenue in terms of increased sales. My point here is that if you have 20 offers on the shelf and only one of them is yours and you take down the 19 others - which one of the offers is the consumer going to pick? They're going to pick yours, right? And through that you will see an increased number of sales, even though you didn't really change anything in terms of your sales and marketing strategy – you just started to clear the market a little bit of all the bogus offerings.

TR: The first one is definitely data analysis. You're able to use technology to analyse a very large amount of data in split seconds, which is not something that a human could do. You could use this to identify trends and patterns, and to create clusters of sellers working under different aliases. Again, this is the talk of breadcrumbs being spread everywhere and you just need to collect the pieces. We also use technology for image recognition, as I mentioned earlier. So now we can go through hundreds of thousands of images in a very short amount of time to look for copyright infringement, which is something that's technology based to its core. Also generating DMCA letters and complaints of notice to take down to websites.

We're also using technology to label findings to make sure that we can actually take them down. If it's a trademark infringement of some sort, then we need to assign a NICE class number to the finding - and how do we do that? If it's not going to be done by hand, it's going to be by the images and the text on the page, using tools like ChatGPT to estimate whether this might fit within the IP portfolio of that particular client. This being done by computer is essentially saving us a lot of time, making our services cheaper, but also allowing humans to do whatever humans are meant to be doing - which is something that computers cannot, like connecting some pieces together that are not obvious to the computer, or creating associations, etc. You can use machine learning AI to identify, analyse and essentially automate a lot of processes and just help companies as well as the end user by making the services more cost effective.

A lot of companies see brand protection as a cost for the company that they are not too willing to invest in, even though in the long run it kind of works like an insurance for you. Now you're able to lower that cost because you can automate so many things. I think this is a great way of using technology – making it cheaper and more affordable for IP owners to actually protect their brand.

There's also anti-counterfeiting technology which is widely used for companies to add some sort of physical security measure to their brand. If you buy a laptop and you open the box either on or in the box, there is a sort of “invisible” QR code that you are not aware of until you scan it. And the counterfeiters can't really replicate that. There might also be some sort of plastic film that you need to pull off which again is something that makes it hard to replicate. Again, this is technology-based but there is digital watermarking of your images online which embed a certain pattern within an image that is something that you can't see as a human but something that works as a sort of invisible QR code as well. And I think those types of technologies are brilliant because they can then be connected to a blockchain as to even further protect your brand and to increase the chance of authenticity. But also when you scan something with that app to verify that it's not a fake, then if it is a fake, that could immediately be reported to the cloud somewhere alongside the geo-location of the app user.

TR: Spotting very, very, very small changes between true and fake or genuine and fake. That's something that a person cannot do, we need technology to do that. Deepfakes are a good example of it. You can make a deepfake and then it's really, really difficult to tell the real and the fake apart - but a computer might be able to let you know instantly because of something that's hidden to the eye. Analysing images for copyright infringement and so on, it comes down to tiny pixel values, and sometimes within those pixels is where you'll find the digital watermarking. You can also essentially say that since what we're doing is within the legal world, you need someone who doesn’t make too many mistakes because those can be very costly for you. And then how do you get an employee that can work for you 24/7 without being tired, always producing the same output? Well, that's a computer. This goes for filing complaints, filling out forms and running searches during the night where, yeah, some people might drink a lot of coffee, but they're still not able to pull the same amount of time as our virtual machines.

RH: Thank you for your time.