Loti Dashboard
In an era where artificial intelligence can clone anyone’s face and voice with frightening accuracy, a new battlefront has emerged in the fight for digital identity protection. Leading this charge is Loti AIa startup that’s tackling the growing crisis of unauthorised AI-generated content and deepfakes head-on.
Luke Arrigoni, CEO at Loti AI
“We’re a licensing and protection platform for public figures, but more broadly, we’re this kind of control critical infrastructure piece for Gen AI,” explains Luke ArrigoniCEO and founder of Loti AI. “If someone makes a fake thing of you, you would want to be able to find it quickly and take it down.”
The landscape of digital impersonation has become increasingly treacherous. From fake celebrity endorsements for non-existent products to unauthorised AI-generated content, the threats are multiplying. What makes these challenges particularly thorny is that traditional regulatory approaches may be insufficient to address them.
“Someone will create a deep fake endorsement of a fly-by-night vitamin company, where they won’t even ship it,” Arrigoni explains. “They just take your order and then they shut it down. They never intended on participating in the market. They always intended on scamming. But you put a famous person on Facebook talking about how this cured their diabetes or something, and now people say, ‘Well, why would this famous person endorse this product? Sure, here’s my credit card.'”
Beyond Celebrity Protection: Locating All Unauthorised Assets of You Online
While Loti AI’s current client base includes artists, athletes, creators, executives, and even high-ranking military personnel, the company’s vision extends beyond protecting public figures. “We do have politicians, and we actually did sign some people that are pretty high in the military,” Arrigoni notes. “We have a very broad swath of what we call public figures, these people might show up in the news for any number of reasons.”
The startup is developing a consumer-based app, acknowledging that identity protection in the age of AI isn’t just a celebrity problem. “We’re working hard to try and figure out what that looks like, because we know that not only public figures, but everyone has this problem to some extent,” Arrigoni says.
A Digital Watchtower for Your Online Presence: You Decide What Stays Online, What Goes Down
The company’s dashboard serves as what Arrigoni calls “a little private NSA of you.” He explains, “It’s all of the deep fakes that show up, and then all of the impersonations, all the real videos. So you have kind of this broad spectrum of how you appear online.”
The system is designed to be automated and efficient. “We have these rule sets, where it says if it is an impersonation account always take it down right away, or it’ll say something like, ‘If a deep fake on YouTube go ahead and issue a takedown there,'” Arrigoni describes. “There’s this kind of control mechanism for people that might have to deal with thousands of them, where it’s not individual. They don’t have to go and click on every single one.”
The Technology Behind the Shield: The Takedown Machine
What sets Loti AI apart is its technological capability, particularly in face recognition. “We do face recognition and voice recognition, deepfake detection, and of course, that auto take down,” Arrigoni explains. “We do this at a level that no one else does.”
The technology’s development has an interesting origin story. “We actually ran a consulting company for 10 years,” Arrigoni reveals. “Two years ago, we had a foreign intelligence agency reach out to us and say, ‘Can you do face recognition on part of the face?’ We’re like, ‘Yeah, we can.’ They delivered the data set which no one else on earth could possibly get. We built the models, and then we retained rights on that.”
The company’s takedown process is methodical and escalating. “If it’s one of those not great sites, we send them a takedown notice,” Arrigoni explains. “If they don’t issue a takedown, we issue a takedown to their domain registrar. If that doesn’t work, we issue a takedown to their host. And if that doesn’t work, we issue a takedown to the credit card processor.”
This approach has proven highly effective. “We can get something taken down within one day 95% of the time,” Arrigoni notes, though he’s quick to clarify their intent: “I don’t want anyone to think I’m bullying some little, tiny, small business. I’m not. They’re usually explicit sites that make an entire business around saying these people don’t deserve privacy.”
Measuring the Problem: a Deepfake Index of the Web
To better understand the scale of AI-generated fake content, Loti AI is developing a “deep fake index.” Arrigoni explains their methodology: “We’re going to admit that no one on earth can actually say this is how many deep fakes were made in a certain month. But we are going to hopefully give a little colour to it by saying, ‘Hey, out of all the musicians we track and all the politicians we track and the CEOs we track, this is what that looks like.'”
The company has established strong partnerships across the entertainment industry. “We have a really strong partnership with WME, but also have good partnerships with UTA and CAA,” Arrigoni says. “We have a good relationship with SAG, with RIAA.”
Their revenue model includes what Arrigoni calls “the offense product.” As he explains, “If you want to put yourself into a contract, we charge 3% of the contract for that. It’s basically a 3% cut on the whole Gen AI market. If you want to put a famous person into something, their lawyer will say, ‘Well, you got to enforce this through Loti,’ just because we don’t trust that you won’t make a crazy version.”
From Bootstrapped Consulting to VC-Backed Product
Unlike many AI startups that launch with significant venture capital backing, Loti AI’s journey began as a bootstrapped consulting company. “If we didn’t earn the dollar, we didn’t get it,” Arrigoni recalls about their early days. “That’s a different kind of Darwinian pressure. You have to be able to deliver – it doesn’t exist if you don’t take the correct deals at the correct time.”
This bootstrap mentality shaped the company’s evolution from a consulting firm with defence sector clients to a product-focused company protecting individual identities. The transition happened gradually: “About a year, year and a half ago, we sent out our last invoice as a consulting company. We switched everyone over to the product company,” Arrigoni explains.
The company’s origins in defence-adjacent technology might raise eyebrows in Silicon Valley, where dual-purpose technologies have historically been viewed with scepticism. However, Arrigoni emphasises that their military consulting background doesn’t define their future direction. “We aren’t trying to seek out military contracts in any way, shape or form,” he states. “At the same time, if you’re a top secret clearance holder, or you need to be protected from internet blackmail, we’re not going to turn people away from that. If there’s a national security reason why we don’t want people to blackmail these folks, just because you’re part of the military doesn’t mean it’s something that we would just say, ‘Oh no, that’s not our thing.'”
The company has recently closed a nearly $7 Million Seed Round, including venture capital backing from firms including FUSE, Bling, Khosla, Ensemble, K5, and Alpha Edison. However, they’re maintaining their disciplined approach to fundraising. “We get lots of incoming requests that we unfortunately have to decline because we are done fundraising,” Arrigoni notes. “I’m only going to raise it if I have something to buy. Right now we have revenue, we have goals, we have a good runway.”
Loti Fights for a Human-Centric Future
Despite their sophisticated technology and intelligence agency origins, Loti AI maintains what they call a “human-centric” approach. “We want our customers to be human,” Arrigoni emphasises. “We want people to think about our product as being synonymous with the other side of AI, instead of AI constantly trying to take people’s jobs and ruin human creativity. This is a piece of technology that is meant to give humans their humanity back.”
As AI-generated content becomes more prevalent and sophisticated, the need for robust identity protection solutions will likely grow. “You only get one face, you only get one name, you only get one voice,” Arrigoni reflects. “Why don’t you go ahead and protect that?” In an era where anyone’s identity can be digitally hijacked, that protection may become as essential as antivirus software was in the early days of the internet.
