Go back

AI, Digital Identity & the Future of Modeling: Spotlite Founders Benjamin Hori and Hannah Choi

24m 8s

AI, Digital Identity & the Future of Modeling: Spotlite Founders Benjamin Hori and Hannah Choi

Spotlight, founded by Ben Horry and Hannah Choi, began as an online marketplace to modernize the modeling industry by connecting models with clients efficiently. Motivated by Ben's frustrations as a model and Hannah's startup expertise, the platform has evolved to tackle emerging AI challenges. After users reported unauthorized use of their likenesses in AI-generated content, Spotlight shifted focus to building a "human IP infrastructure." This includes developing tools for IP protection, launching in March, which scans the internet to flag unauthorized images, and a data licensing system to ensure models are compensated when their likeness is used in AI training or campaigns. The founders emphasize ethical consent and ownership in the digital age, aiming to protect individuals' identities globally. With plans to expand into the U.S. market in 2024, Spotlight seeks to set a standard for transparency and fairness, balancing innovation with the need for digital rights and authenticity in industries like fashion and entertainment.

Transcription

3992 Words, 21311 Characters

English
Hi, Melissa Bell and welcome to the business roundup where we catch up with the movers and shakers in the business world. Today joining me is Ben Horry and Hannah Choi. They are founders of Spotlight. Thank you so much for joining me today. Good morning. Nice to meet you. Now Spotlight is an online marketplace for booking models and creatives based out of South Korea, correct? Yes. Okay. So before we kind of dive into this because I find this like really fascinating, especially when you have like Hollywood actors and different things that are going on with like like Mrs. and AI, I think this is a pretty cool topic. So before we dive into what you do, do you mind kind of talking a little bit about your background and what kind of prompted you to found Spotlight? Definitely. So my name is Benjamin. I was in the fashion industry for, wow, a really, really long time. I like over a third of my life. Primarily working as a model. I had the opportunity to work North America, South America, Asia, just about every continent except the polls. And during that time towards the end of my career, I just got really frustrated with the lack of digitalization in the industry. It was really, really tough to find efficient ways to connect with clients or get visibility on my schedule, my payments, everything. And that kind of led to the early foundation of Spotlight becoming a simple marketplace very much like Airbnb is for you to go and find a house. And then through that journey in a very early stage, I met Hannah, who has kind of been the powerhouse to really carry this company to where it is now. That's awesome. What about you, Hannah? So I'm Hannah, very nice to meet you all. My whole career was in startup. So my first company was Clark Kitchen, which was Uber Founders at Constartov, which was running a lot of Ghost Kitchen for the delivery system. So my first job I saw from lunching to all the way to really fast growth and then expanding this business globally, which was fascinating how startup solving the actual problem. And then I went to the Uniquan startup that's excited to match group in Korea. And then I thought, well, Korean based startup can also scale. So ever since then, my whole career was in that ecosystem. When I met Ben, I was in the early stage VC. So because I really love startup ecosystem, I decided to meet more founders and then be helpful and try to invest in the early stage. And I was reviewing a lot of deck at that time, like hundreds and hundreds of them. And Ben approached to me and, hey, Hannah, there was a real problem in my industry that I really want to solve. And this is an idea. Teach to me at first, as I know, because I saw how hard it is to be a founder as an investor. So there's no way that I can get it. More and more we talk about it and we figured that we could be a good team because he said industry and domain experts. I know a little bit about the startup. So we teamed up and three years later, here we are. Yeah. That's awesome. And now for people who don't know, so I know it's, so what you're building is what they call a human IP infrastructure. Now first, before we kind of get into that, what kind of is the background on that? Like so, because we have AI and there's likenesses. Can you walk us through that a little bit? Sure. Yeah. Okay. Okay. So it started from actually our users feedback and then chat because AI really started to disrupt being this industry after we started this marketplace. And some of the users and our close friends, they started to send us some messages that, hey, I did not agree to this shoot. And then my face has been used in to this a lot of different ads that I see on Instagram, like what's happening. I never wore those clothes where I never posed out way, but it's using there. So at first, we thought that was a big problem. And then we dig a little bit more how those models being trained to generate those image. And it turned out there are a lot of settlement and loss that going on about those tech company has been scraping out those human likeness identity and imaginary from the web. And then trained actual foundation model that are used for commercialized product, which is making a lot of genetic human. So we thought, okay, first of all, our likeness has been infringed upon because I mean, with those gross of the AI. And second, if the foundation data has been stolen from our big user base, we have to be compensated by it and then have the ownership back. So it's a little bit of shift from modeling marketplace into the very deep tech AI space. But at the same time, the shift was pretty organic because it was the user's problems are getting bigger because of the market shift. And then we thought that's something there's something we should do about it as well. So that's where we started to diving into it. And now, where do you think now, or what's your goal as far as like maybe five years down the road? Where do you want this to kind of go? Because I know they were just talking about, I mean, this might not be the same thing, but I think it was Brad Pitt and somebody else. It was like a movie scene that was not really them. And that was like a big thing that, you know, it was like a big Hollywood situation. What's your goal for like in the next five years? Of course, I would like to start from next three months first because we're building something very exciting. First of all, we're launching something called IP protection. And what it does is when you upload a photo of yourself, we search the internet and then flag some photos and video that looks like you and then let you know that hey, is this you or is it is you, but have you agreed to this shoot or is it a Gen AI image? Happened without your consent. So we try to get user have visibility and then alerted whenever those infringement is happened because now is very easy to just remix your face like it's a product which is not okay. And there are a loss and policies that actually prohibit that action as well. So we try to make user aware of what's happening and then help them take action as well. So when you find something that you want to take an action, our service provider actual report that is reviewed by our legal councils, then you can download it and then take an action immediately. So that's something we're working on that is launching in late March. So we're really harmed that. And second one is the one related to the foundation model data training. So if our data was used, especially if your actresses model your image are out there like so so so often and high quality, you are exposed to those great being for sure. So what we try to do is okay, we understand there is a needs and then our data is valuable for them. And then why don't we take the ownership back so that you can license your likeness for either for getting the right compensation or even your likeness can be work in the new campaign that when you agree to it. So you can get the right compensation. So that's called the data licensing. So that's second one we're working on. We're just doing the POC with one of the biggest buyers in the industry. So those are what we're working on in next three months, which is very exciting. And in five years, I think our hope is our likeness, which is our face need to be treated the right way because this is something very valuable and unique to you. And it should be protected. And if the company is trying to use that data to get, you know, like commercialize it, you should be the owner who allows it. So we want to build that huge infrastructure whenever the face data or your likes data is used either for the training either for the ads. The owner should be allowed the visibility or agree to it and get compensated. So that's how we're picturing it. Yeah. That's yeah, that's super important because not only I mean, you want to have this creative control, you know, back in as far as like, you know, agencies used to at least make sure if there was like an image in a magazine, it's a little bit easier because they're you know, someone's going to see it and they're like, I didn't agree to this, you know, I didn't sign the papers on this, but when you're dealing with the, you know, online and the internet, it's so vast. So having something like that that can detect this is amazing. Well, fingers crossed. Now, now what about you now what what has been your experience then have you had as far as in the industry? Have you come into like, you know, have any issues in this type of situation where you've had like your likeness taken? Yeah, I mean, it's really tough to say because the products that we're building now don't really exist to the extent or scale that we want them to. So you might be able to go to Google and try to reverse image search something like your face, but you're not really going to get the kind of results or feedback or visibility on where your face might actually be used, even if it's been slightly manipulative. related. So it's really hard to say. But the catalyst for actually building these products now was I have quite a few friends that are still active in the modeling industry and then talking to them over the last few months. They just started sharing all these instances where clients were repurposing them or using AI tools to just completely create new versions of them, like changing their ethnicity, but not changing anything else in the photo. So it's like this really weird ethical line we're dealing with right now where everybody deems data on the internet, at least public data as fair use. And I think regulators really need to start like taking action to protect people's digital identity because you know there's so many cases where this gets abused. There's deep fake, there's explicit content and you know this is like just a surface. And I'm sure if somebody found an image of their face or their family member or their friend or even their child or something being exposed or manipulated in some kind of crude manner, it would be a big problem. Absolutely. So I mean like right now I mean this is more for you know for the creative industry, but do you see maybe it being beneficial for just other you know across the board as far as you know being implemented in different industries and just on the internet in general. Of course I mean everybody's digital identity needs to be protected and then they need a way of providing consent or opt-in or opt-out. You know we've been so quick to adopt AI technology in the first couple years of this boom where nobody was really actively thinking about protection layers where we just wanted to build things quick everybody wanted to be the first one and then safety and protection kind of came as an afterthought. And now it's starting to get to the point where it might be a little bit too late, but I do believe we're like right at that cost of building in these protection layers for everyone. So I mean there's you probably use social media right Instagram, Facebook, TikTok, X. So I don't know if you know, but a lot of these terms of service policies for for major SNS companies, they get your data for free. So if you upload content onto Instagram they can repurpose it for AI model training. I don't think that's necessarily right. I mean it is their service, but some form of compensation needs to be paid or at least some visibility or opt-in and opt-out terms. Usually those terms are pre-restrictive for some users. And we're just trying to change that. That makes sense because like you're saying as AI is transforming all industries, I mean kind of that line between innovation and exploitation is that a B, we have to have a better definition there. And now what does trust actually mean in the world of like deep fake and digital twins and that type of thing? I now deep fake I do know what is digital twins and then there's synthetic media. I have not heard of those. Do you mind kind of talking a little bit about those? Should I mean they're all kind of the same thing. So deep fake or digital twin or synthetic media is just anything that was created using a generative AI tools and is not actually authentic. So let's say I went to use a popular tool like Nano Banana, used myselfy and then maybe somebody else is posing. So Hannah's photo and then I just embed my face over top. So that would be like in-face painting. The same kind of tools kind of exist for video. But our goal is to not necessarily remove innovative tools that allow for faster production or more efficient production. But we still want to find a place for real people to exist in you know this new generation of synthetic production which is doing everything with generative AI tools. So yeah adding on to that digital twin will be your digital self that you have ownership to it and then you can always control how this likeness will be used. So for example your digital twin can be in beauty, campaigning Korea. But then you will know where it's used and then you provide consent and then you get compensated by licensing it. For example you can use your own digital avatar to make your UGC campaign but then it should be yourself or your Asian has exclusive right to it allow it and then know the scope and how it's going to use. Like how we've been building this for very long time in modeling industry. Yeah and you know that some good things that you were talking about too with like Spotlight. So one of the other things that you do you were which I thought it was interesting. Let's say you're you know a model or an actor or you're somebody who is in another country but you want to do a booking you know you can do that. Like how could they set that up like in the future now you could be in two places literally at once and probably be able to do a shoot. Yeah I mean I think that's something exciting about AI era once this consent infrastructure or how the digital likeness is used as standardized. So right now technology is there is very easy to screen capture someone's face put your digital put your product and then make the pose at 3D or any way that you want and then make a perfect looking digital image is already exists and then a lot of services are doing that. But what is missing is the permission from the actual person. So if those things are connected what we're imagining right now is Spotlight at the moment you get the job request from within your area or internationally you have to be on the shoot on site. But what we're thinking of a next step is if any advertiser or brand really like your likeness like like your aesthetic they can just send your request for just licensing your likeness and if you agree to okay this will be not for you sir ABC that I don't agree to and the usage limit is going to be this or that the length and everything if you agree to it that we can provide the way to use those users face and then make the genie image and then the brand will have risk free because you're not really you know breaking anyone's life and then the user can actually be appeared in more area or more regional basis than ever before. So I think one thing how do I say like once there is a way existing for the face owner to be agreed to those campaign I think this is a new exciting um how do I say chapter for the modeling too yeah yeah definitely because again you'd have to like let's say I mean there's there is a lot of like people going from different you know whether go to Japan or going to New York or whatever but even more like I would like it'd be like you have more access to different types of ethnicities it would be kind of exciting because you don't have to if you have to apply you know 16 hours that's a little bit of a trip but if it's just really quick and you can do that's it's it's it's pretty cool but like you're saying so you basically you need to be able to kind of take the ethics and kind of put this into the product development before we can kind of move forward in these kind of areas definitely and I think there will be more and more brands who value that as well because authenticity has been a big part of any brand especially in fashion and beauty as well so having that thought into your product or marketing that we are using not just a fake faces but um likeness that has a consent into it and then using that ethical way of those AI usage in your brand I think that's something can be plus to your brand as well. Very true and now you're scaling across Asia and the US um what why guess this is one of the questions but what are some of the things that you're noticing as far as like differences like cultural differences or things that like like that you've kind of come across to. I mean I think I think all the markets kind of operate the same at least for this niche of the domain but you know entertainment sectors pretty broad you actor actress influencers commercial models faster models run way models like there's a really wide net here but from a client's perspective at the end of the day they just want the face that works for their brand no matter what the job is and I think that's kind of the same in any market so at least in Asia I think people have been pretty quick to adopt using generative AI tools to start creating new content um just because it's cheaper and faster but then again there's a huge underbelly of issues there that nobody's really talking about especially when you know these brands start to produce this content and release this content the backlash is usually pretty severe when people can tell that it's generative AI and I think the same goes for the US people don't really like to see what is like the coin term now is like AI slop or you know just really poorly created AI content. But I think from a client perspective, they are looking for faster and cheaper solutions to get their products out there. And it makes a big deal for emerging brands or designers just starting out or brands that want to cast a wider net using UGC contents and different languages with different ass methodies, shapes and sizes. So I think everybody's looking for the same new tools, but everybody still also has the same worries of what happens when they release those kind of contents. Like for people who want more information, would they go to the website? Now do you take on like additional talents and creative at this point or OK? Yeah, we're now the largest platform in Asia. I would say arguably we're the largest professional platform worldwide. So we're always taking in new talent. We're always looking to work with new agencies. Our goal is really just to provide accessibility and then more exposure to different markets. And we're trying to do that in the most ethical and safe and fairway and transparent way that we can. And also like are your plans to expand more in the US as well? Yes. That's our this year's biggest plan. And actually the differences between Korea and US, I think the most important thing is compliance. So now we are getting legal review and then how we can make the best terms and condition and the service flow for the US user. So we're revising and tweaking descendads. So once that's ready, first we are releasing the IP protection where you can see where your likeness uses used by end of March. And then by end of June or July, we're launching our marketplace service as well in US where you can get the job opportunity in person shoot within your area or it could be direct to Asia as well because we have a global client base. So that's something we are very, very excited about. And because now you're actively building, if there's anyone who would like to test it first or give us opinion on how we should build it, we're super open. Ben's been talking to a lot of management agency based in US. So that makes sure we build something they want, not only Korean client one, right? So yeah, we're very active. I think a lot of people are looking forward to Korea and US in different places. Well, it's exciting. And so people want to kind of keep updated again, the website when you make the announcement in March, they can find it there. Yes, for sure. So we have a website, a home.spotlight.global. There are a lot of updates and news going up there. Also, we have an Instagram as well. And we try to do the founder-led content and then build in public. So actually our users can see our product updates. While we did this week to make those issues sold or what are the tasks that we're having, user interview, and everything was live broadcasted on the Instagram. So if anyone's interested, definitely follow us. And then comment a lot so we can build something that you want. I thought it's been very great speaking with both of you. And I think it's really exciting. Because like I said, something that I think is lacking. And I think it'll be great for the industry if we kind of take these initiatives now. Well, thank you so much. Yes, thank you for having us. Thank you for joining the business round up where we always try to bring you interviews with some of the leading people in business, whether it be technology, finance, or energy. I hope you tune in. I hope you subscribe and listen. And we'll give you updates for new interviews that are coming up. Until next time, I'm Melissa Bell.

Podcast Summary

Key Points:

  1. Spotlight is an online marketplace for booking models and creatives, founded by Ben Horry and Hannah Choi, initially to digitalize the inefficient modeling industry.
  2. The company is pivoting towards building a "human IP infrastructure" to address AI-driven issues, such as unauthorized use of likenesses in generative AI models and synthetic media.
  3. Key upcoming features include an IP protection tool (launching March) to detect unauthorized image use and a data licensing system to ensure consent and compensation for likeness usage.
  4. The founders aim to create a global standard where individuals own and control their digital identity, enabling ethical use of likenesses across industries.
  5. Spotlight plans to expand its services to the U.S. market in 2024, adapting for compliance and cultural differences while maintaining a focus on transparency and ethical practices.

Summary:

Spotlight, founded by Ben Horry and Hannah Choi, began as an online marketplace to modernize the modeling industry by connecting models with clients efficiently. Motivated by Ben's frustrations as a model and Hannah's startup expertise, the platform has evolved to tackle emerging AI challenges. " This includes developing tools for IP protection, launching in March, which scans the internet to flag unauthorized images, and a data licensing system to ensure models are compensated when their likeness is used in AI training or campaigns.

The founders emphasize ethical consent and ownership in the digital age, aiming to protect individuals' identities globally. S. market in 2024, Spotlight seeks to set a standard for transparency and fairness, balancing innovation with the need for digital rights and authenticity in industries like fashion and entertainment.

FAQs

Spotlight is an online marketplace for booking models and creatives, originally based in South Korea. It has evolved into a human IP infrastructure platform that helps protect digital likeness and manage consent for AI-generated content.

Ben Horry, a former model, was frustrated by the lack of digitalization in the fashion industry, making it hard to connect with clients or manage schedules and payments. Hannah Choi, with a startup background, joined to help scale the solution after recognizing the problem's potential.

Spotlight offers IP protection tools that scan the internet to flag unauthorized uses of a person's image, including AI-generated content. It also works on data licensing to ensure users can control and get compensated for their likeness in AI training and commercial campaigns.

In late March, Spotlight is launching an IP protection service to alert users about likeness infringements. By mid-2024, it plans to expand its marketplace to the U.S., allowing job bookings and ethical AI licensing for digital likeness.

With the rise of AI, unauthorized use of faces in deepfakes, ads, or synthetic media can exploit individuals without consent. Spotlight aims to give people ownership, visibility, and compensation for their digital identity to prevent abuse.

Spotlight is focusing on U.S. expansion in 2024, adapting its services for compliance and user needs. It already operates as a major platform in Asia and aims to connect talent with global clients through ethical and transparent processes.

Chat with AI

Loading...

Pro features

Go deeper with this episode

Unlock creator-grade tools that turn any transcript into show notes and subtitle files.