Birth certificates for AI agents – DNW Podcast #587

Welcome to the Domain Name Wire podcast. I'm your host Andrew Olamin, and this is episode number 587. My guess this week is Akram Atala. He is the CEO of Identity Digital. Identity Digital is the largest top-level domain operator when measured by the number of TLDs. It's the company behind.live.bet.bio and hundreds more. It also operates the popular.ai and.io namespaces. The company recently announced DNS ID, which it builds as birth certificates for agents. Today, Akram explains what it's building and what it hopes to achieve. We also talk about this year's new top-level domain name expansion. So stay tuned for that interview in just a few minutes. First word from our sponsor, our sponsor this week is Spaceship, where you can build and manage your entire digital world. Spaceship offers a huge catalog of top-level domains for some of the lowest prices around. One way to find domains quickly is to use its beast mode search option. It lets you compare up to 5,000 keywords at once, narrow your search with advanced filter options and bring intelligent suggestions to the service. You can choose to include domain hacks, to pluralize nouns, and drop the last vowel to find hidden gems. Right now you can get a.com domain for just $5.67 with the promo code com67 that's limited to 1 per account and.ai domains are just $79.90/year with a 2-year registration. Head over to Spaceship.com and don't forget to hit the beast mode button when you search. Let's discuss some of the news in the domain business from the past week. First up, Internet Commerce Association announced two changes to its board. Alan Schifflett of GoDaddy is joining the board. He'll be taking over the seat previously helped by James Bladall who is retiring from GoDaddy. Alan runs GoDaddy's aftermarket business and will be a welcome addition to the board, while James was only on the board for a short time. I'm sure I.C.A. would like to find another board member with a policy background like his. In Jeremiah Johnston is leaving his role as General Counsel at Ceto and will also step down from the board. Jeremiah and Ceto have been there from the beginning and they've seen I.C.A. grow against the odds. Dish DBS launched the.latino domain name in sunrise last week. The company previously launched.moble and it still hasn't launched a number of domains including.data and.dot. I would expect them to try to roll them out before new domains from the current top level domain expansion round hit the Internet. I.C.A.M. published its 2025 tax return last week covering the 12 months ending June 2025. The return shows that revenue was down about $10 million in a prior tax year with about half of the drop attributable to lower program service revenue and the other half to lower investment income. On a tax basis, I.C.A.M. lost about $12 million during that year. The return also allows you to peek into how much top employees are making including other compensation. Nine people made a half million dollars and more. There's a link to the return and my post on May 11th. Also last week, I wrote a warning for businesses to not let their web developer register or control their domain name. I cited a recent lawsuit with a company suing its developer who they say is holding its domain hostage after it decided it wanted to switch development firms. I see this sort of thing all of the time and it's a shame that companies go through this. Again, the domain names are not top of mind for most business owners, but they are too critical to give the keys to a third party. Head over to domainnamewire.com for more details on these stories and to read other stories I posted last week, including a lawsuit over a.bett domain name and a list of 33 companies that have bought.ai domain names on the aftermarket this year. Now let's chat with Akram. Akram, welcome to the show. Thank you Andrew. It's good to be here and I noticed that you asked your audience who they want to see on the podcast and I wasn't mentioned but here I am. Oh, we can give them something that's interesting. I'm sure they'll enjoy listening to this show. So you recently announced a new initiative called DNS ID and this is under the umbrella of Innovation Labs. First of all, what is Innovation Labs? I'm going to give you the short answer but I can give you the long answer if you want. We thought that basically with the agents that are coming online, we have an opportunity to position the DNS as the best solution for identities for agents because as you know today agents mostly are working within the enterprise but then they're starting to cross outside of the enterprise and go across the internet and all of that. And there is nothing more open than the DNS for these kind of solutions. And as we all know when blockchain technology came about, our industry sat on the sideline and watched it develop solutions and eventually we didn't do anything and that didn't work and now they're converging back on the DNS. So we at Identity Digital felt like this is an opportunity we should not sit on the sideline and wait for things to develop. We should actually show what the DNS capabilities are and how they can solve these problems using the DNS which is the technology that's been proven for so many years. It has governance. It's an open standard. It's available for everybody. Accessible across the globe. So we believe that this is the right solution for that and we found innovation labs to make sure we address this and we work with the industry on a solution that it can be used for addressing that need without disrupting our core business. And so this is really why we created this division separately. Okay. And so innovation labs is solely focused on DNS ID or does it plan to look at other possible innovations for using the DNS. So for now that only focused on DNS ID and hopefully we successfully that and we can build the standard around that and then after that we'll see if there are other opportunities. I'm sure we can deploy the resources to do more of your things. Okay. So DNS ID you kind of talked about that kind of at a high level here and answering that question which is great. This is a way but like how would you talk about it? What does it do for agents? How does it solve this problem that we might experience here? So I think that the problem is big. You know an agent that is working for a retailer needs certain verifications and identification that is different than an agent working in a bank for example. So the problem is not a single solution. I think that the problem is big and depends on the functionality needed and the security and verification and authenticity that's needed for the solution. But what we're trying to do is to establish the DNS as a core function in that solution. So we think that basically by providing an identity for the agents that is immutable, that is very viable, that is a durable and persistent is the core of the need. And then a lot of different things can be built on top of that as you know as multiple layers and other than security layers and other things can add to provide a commercial or customizable solution for different needs. Okay. And so the overall idea here is adding kind of credibility to an agent. Like can I trust this agent or is this some rogue agent that someone created that is now being put to use that? Well, it's more like an identity. So I think the team talks about it as a birth certificate. Your birth certificate is needed to get you a passport, to get you a rival license, to get you. It's a core that is who you are. And from there you get different solution and you address different needs, right? And so the same thing, the NSID is trying to provide who the agent is, who's responsible for the agent if you want. And then it's providing a ledger in the blockchain that is always available so that you can look at as agents evolve and change and do different things. You know, all of that of those records are kept there so that you can always look them up and understand what the agent has gone through and whether it's been successful or whether it's all these things are all in drag. Got it. So when I look at that, you're not the only company trying to do something in this space. And the IETF, the Internet Engineering Task Force has a standard out there, draft standard I should state. And GoDaddy has actually built something called the agent name service based on this draft standard. So is your solution also based on that or how does it differ? So we're working with GoDaddy and other industry providers to make that standard work for everybody. And the core, you know, our standard says basically that you use the DNS, you use a PKI and you use blockchain to actually provide the DNS ID solution. Now we're hoping that the standard will be a common standard and DNS ID will be one implementation of that standard. That's what we're working toward because we don't want to fragment this. We want to actually come at it with a single solution for the whole industry to be behind. - So if I look at, I wanna make sure I understand this. So GoDaddy has kind of done an implementation of the IETF draft standard and you are creating kind of an implementation based on the idea of using the DNS here. But wouldn't these kind of compete? Like if I'm creating an agent, would I choose DNS ID over GoDaddy's agent name service? - Well, I think that the agent name service of GoDaddy goes up to the layers above what we're calling the standard while we're sticking to the standard. Think of the DNS ID as an implementation of the standard while DNS is an implementation of the standard plus additional function. - Okay. - But you're using blockchain, which I noticed that GoDaddy isn't using as a technology and the IETF draft mentions that, "Hey, you know, blockchain has some limitations that we're not sure if it would actually work for this." And again, this is a draft, right? So this is kind of an ongoing thing. But I say, you know, there's high latency, skill ability issues, that sort of thing. So what do you think blockchain is key to have as part of this? - Because it's durable and it's global now, everybody can access the blockchain data if they have the right key. So overall, I think it meets our needs. If then is it better solution, we will be looking at that. But I think at this time, we believe that blockchain is probably the best solution for keeping records and the data of the agent available, secure, and reachable. - Because if I think of registry technology, right? It's the registry manages this database or this ledger, but it's not distributed. So you're saying for this agent system, you think that blockchain and having this public record makes more sense. - Yeah, I think that we're trying to build our implementation on the open standards and be able to provide the solution to anybody out there that wants to verify what the agent is. And if it's proprietary, then it fails in that openness that we're trying to do. So that's why we think it's the best outcome. And we're actually looking at it from a perspective of even open source. We don't want to own this and say, this is ours. We want to actually make the solution and make it available for everybody, even make the source code available for anybody that wants to implement it as well. - Is the idea then that much like I can choose between.guru and identity digital domain and.com, which Verrissime manages is the idea that when I create an agent, I can go to any one who is built on this standard to get my quote unquote birth certificate. - Yeah, I mean, the idea is that it's an open standard, anybody can implement it and what you use, which blockchain you use, which the idea you use, that's up for the implementation. - Got it. Okay, interesting. So then, because one of my concerns is, okay, identity digital, big player in the domain space, go to any big player in the domain and website space, but what happens if Google says, okay, here's our system, that's sort of thing. And so the goal would be that anyone who builds a system like this, whether it be Google or anyone else, open AI, would base on this standard, this agent name service standard from the IETF. - Yeah, I mean, the idea is that really, it's not about identity digital loan, it's about the DNS actually. So it's that the DNS is the right technology to use for an open standard, open approach. And we're trying to enable that and show how it can be done. And then hopefully there'll be a lot of solutions, build on top of what we provide and so on. We're not in this for something just specific for identity digital. We're in this to make sure that we point to the DNS and we make the DNS a default place for solving these identity issues. - So the benefit then to identity digital is if people adopt this, more people register domain names. And hopefully some of those will be yours. - Yeah, I mean, at the end of the day, the more people use the DNS, the more they need a registry and the more they need domains. And so it fits in our space. But also it's an learning from what happened to the blockchain if you remember. So right now the blockchain, a lot of the blockchain solutions that were out there, trying to do their own registries and their own solutions for being able to access and be accessible on the internet, didn't work. And now they're coming to the new HDLD program to actually participate in that and solve that problem. And we could have solved that problem back in the day when blockchain launched if we had worked with that on the solution instead of sitting on the sideline as an industry. - Got it. So like finding a way to integrate Web3 technology with DNS rather than Web3 is separate from the DNS kind of approach. And ultimately as you mentioned earlier, they are coming back to the fold, realizing the limitations of blockchain technology and alternate naming systems, which those of us like you and me have been in this industry for a while have seen these alternate routes pop up from time to time and they always face significant challenges. - Yes, that's the appropriate thing. I mean, the minute that you go - Properly it did it, it's just not opening. - Yeah, the beauty of DNS is if I type in a domain name, I know I'm going to that domain name. I don't need to say, oh, I'm on this blockchain or I'm on this platform or I'm using this plug-in or this browser, I can use Chrome Safari, whatever I type in identity.digital, I go to your website. How far along is this? Is it just kind of an idea at this point if you started technical implementation, talk to me about that? - We are developing the technology right now. We have some, I would say, alpha product that we can share with enterprises who are working on alpha product for the retail segment also. We're probably, I would say, six months away from a product that we can actually show. - Okay, and so is the idea to offer this through the channel as like add-on, if you will? - Yeah, I think that the big market is probably the enterprise where everybody saying every enterprise is going to have millions of agents, so on. We need to address that and we need to address a commercial structure of how that works and all of this. And then on the retail side, I think that, you know, I don't want to get into speculation of how agents will develop and will become, but we can see that agents could become brands. That you need a brand of domain and that brand of domain is an agent and people go there to transact and do things. And so I don't think of it as any different than a feature just like me, tacked on email to domain name, like we tacked on the website to the email name. And now, you know, we're adding the identity and the agent I let it do it. - That's interesting what you said about brands because I talk to some people in the space and they say, yes, agents will need a domain name, but it doesn't really matter what the domain name is, right? It could be a string of numbers because it's just a computer talking to a computer. It sounds like what you're saying is there will be that, but then there will also be these cases where the brand and the domain name really does matter. - Yeah, I mean, people forget that, yes, everybody says today that we are going to a place where you go to agents and you get your answers instead of going through the work of what the agent did to give you your answer. And that's true, but in a lot of the cases where our interaction happens, you know, human interaction happens with the internet, we like certain things and I don't want to go every time to an agent and they tell me, oh yeah, there are all these other things, but not the thing that I want. So maybe in the discovery phase, I'm looking with the agent and the agent telling me, but at some point, I might like certain places where I want to go directly and when you want to go directly to them, you're going to need the brand, you're going to need it, the main you're going to need an address to go to. And you can see that today, I mean, everybody on social media tells you the minute that you see their content, you might never see that content again unless if you follow me, well, why is that? It's because the engines and the algorithm might not surface that content again. And so if you like that content, you need to take an action and that action is following somebody on social media. Versus on the internet is to actually remember that domain name because that's how you actually go mad. So we're seeing kind of an interesting thing in the domain space. I'm sure identity digital is no different. Q one of this year, like new registrations went up. And I think part of that is because anyone can develop a website now or a SaaS business very easily on a lot of these platforms. In fact, name.com, which you guys own, a registrar has created an API that's working with some of these platforms like Repilit. And so now it's easy, like even I can't code, but I'm spinning up apps, right? And then sometimes I needed a domain for them, sometimes they're internal. But so we're seeing kind of that spike here. At the same time, we're seeing that certain types of content-based websites, whether they're travel blogs, comparison sites, that sort of thing. Traffic to these, This is going down significantly because people are either getting the answer at the top of Google in the AI or the AI snippet or they're going directly to GPD or Claude or something else to get that answer. So how do you see this play out? I think a lot of my domain names that I own and some of them are definitely geared for more of that content site and some are more for a tech or a SaaS platform. I didn't need the digital obviously with a few hundred strings here has some that probably go on each bucket. How do you see that playing out in the long run? I think that that continued to be both kinds of domains. Right now I think I saw a chart from Apollo showing that the business creations is at all time high, which is persistent, what you're saying. I think that we've seen these trends in the domain name business when there is a industry or a economical decline. You see a new, lots of new businesses gets fun out and we see a lot of restrictions because of that. So I think that this is probably a similar time we're going through right now is that there are some layoffs. You're seeing people starting to spin out new businesses and stuff. And so we're seeing some demand for domains. The high tide is lifting all boats right now in the industry. I agree that agents that are working behind the scenes are probably not going to need an identity that is branded and memorable. So we need to figure out how to make that happen commercially. But there would be new brands that will continue to surface and be used by humans. And these will continue to use branded domains. At the same time, I think that the story here is really about the increase across both settings. As both segments keep growing, we're going to need a lot more domains and we're going to need a lot more TLDs and so on and therefore you see the new GTI program coming up now and there is a lot of hype about it and people are expecting a lot of applications and things like that. I think that it's good for the industry, it's good for all of us, you know, what's going on. I don't see that one thing is going to take over the other and it's going to continue to be humans that add content to the internet that AI is actually dissecting, providing as answers. It's a fair point, right? How do the AI systems suggest new content if there's no new content being created? I mean, it sounds like what you're saying. There are a couple of tailwinds here, right? They often start their own business. There's also the fact that anyone can easily create a complicated business right now and then maybe we have this slight headwind in the long run. Maybe media publishers in particular can't monetize our content because they aren't getting as much traffic, but that's obviously something we'll see kind of play out over time. Yeah, I think that you can hear already that some people are looking at ways to have agents pay content providers for creating their content and for using that content. So I think eventually, just like we normalize the humans creating content and getting content and using it and paying for it, I think that will also different models themselves advertising maybe it's a different model, but there will be somehow otherwise you won't have content. If you're not motivated, people, they won't add content. So yeah. Cloud Flare is one of the companies that's doing something there if anyone wants to look more into that. So you said the solution might be out in about six months or might be something to really show here. What does success look like for you? Like 12 months from now are there metrics or is there some sort of qualitative look at what success it'd be? For me, the most important metric is really that everybody comes around and realizes that the DNS is a right solution for identities. Whatever the solutions that actually or the application on top of the standard that we all agree on comes out and becomes more successful, I'm happy for that. I just want to make sure that we're not going around the DNS and then regretting why we did stock with the DNS and the first person. Right. So for you, the metric, at least at this point, isn't necessarily how many birth certificates were created on DNS ID, but it's just that people are coalescing around, hey, the DNS is the solution for this issue. Yeah. Absolutely. Cut it. Let's talk a little bit about this new top level domain name around that we're going through right now. Applications are open right now. So people are getting nose in. I'm sure you're busy working on those as well. This time this process happened at ICANN. You were on the other side, right? You were working at ICANN. This time you're on the applicant side, which is more fun, which is more interesting. I think this is more interesting. ICANN, like we were all doing something for the first time and launching a big application round and having a lot of first times on everything. A lot of things that didn't work, a lot of things that mistakes that happened and it was very stressful time. And we weren't expecting as high a demand as we saw. And so it was actually pretty challenging and pretty stressful. Eventually, we got our act together. But there were still a lot of things that were regrettable. I think somebody applying for a hundred domains and having to test every domain by itself. I mean, that was like silly and everybody looked as funny. But it was like a lot of it was we have one contract for application. And so you have to do each application separately. And so how do you do that and separate all of these things? So I think that we matured a bit like the way that they're doing now, pre-approved, technology providers. So there are like this round is a little bit smarter. You know, we've been incorporated some of the learnings of the past round into this one. So things are better. Of course, there's still a lot of things that could be improved. And I think on sitting on this side, I don't have to worry about a lot of things. I just have to worry about applying and, you know, what strings am I going after? I guess there is an additional thing since you're in the CEO role here, right? You're seeing all of identity digital. There are two sides to the business now because you acquire the affiliates, the registry back end, which already had clients on it. Now I assume in this round or I know in this round you're going for clients on that side, right? Who want to use the back end registry services. And then you're also applying for strings yourself. I'm curious, how are you keeping those separate? You're not the only one in this boat. You know, I assume GoDaddy will apply for some strings and are also a back end provider. How do you keep those separate so there's not a conflict of interest? So we have two separate teams. We have the rest three teams that manage our own and operated TLDs and we have the RSP team that's more of a account management team that is managing our relationships with our customers. And what we do is we are actually asking our customers who want to apply for domains, not to tell us what they want to apply for. That's our first layer of defense. And then we have a separation of duties and we don't share like the people working on what we're going to apply for and not the same people who are working with our clients helping them apply for their own domains. The first thing is don't tell us the string right now. Let's actually just make sure we get through your application going and everything. And then eventually you can add the string and you can apply for it right before and of the application run. I guess I mean technically someone could tell you after they submit the application what the string is or just like, Hey, we want to contract with you to run the back end for five strings. Now, some people need hand holding through the process of the application and such. But for some of the sophisticated players, they could just say, you know, negotiate whatever that price is or contract. I don't know how it's working this time and then reveal the string to you and then hopefully after they apply. Yeah. I think that from our perspective, there are a lot of people out there that are advertising what their application and their string is. So for these, you know, we have to make sure that they understand that we might be applying for the same string. We are very upfront about that and we tell them, look, you know, we don't know what our list is yet. We haven't finalized it. But we're going to be applying for a few of these and so that is a competition there that you know, we have to compete. Yeah. That makes sense. And being upfront about that, everyone understands. I think that's key. Give me your pitch for why a new TLD applicant should use the identity digital back end. Well, I think that our back end is the most modern back end. It was designed for the new digital days. And so it has a lot of the contractual obligations all designed into the solution. We are a definitely not the cheapest. I want to advertise it out upfront. We are not providing just basic functionality. We actually provide a lot more functionality than anybody else because we run our own TLD and we need that functionality. So we have a lot of features now on our back end that, you know, others don't provide from the ability to pick up catch, you know, and provide a aftermarket solutions for your TLD to the normal day-to-day operations. So we build the early access phase as you launch, right? Yeah. We have also solutions for mitigating. of your names premium domain. So you have premium domains, you have, you know, so we have a lot of features that we already include in our solutions that, you know, others don't have and they don't provide you with that. So it's up to you if you want to get just bare minimum stuff and pay for that, that's fine. And if you want to actually get a better solution that helps you market your TLD and benefit from our access to all the restars out there, we also connect to restars that are not I can't accredited for example. So we have probably the broadest restriar numbers out there because we have CCT and ease that we operate with. So from that perspective, we have the more reach than anybody else. So distribution, which I think will be key this time around. And of course, you can't guarantee distribution with a certain registrar. I had Ray King on recently talking about kind of how registrar's work with that. And we might have some slotting fees. That's sort of thing this time around. But I will say, as Ray King said, as a registrar since he runs pork bun, if it's easy, if I'm already integrated with this system, then it certainly requires less of me as a registrar. As you mentioned earlier, name.com. That's really why we have name.com. We're not in the restaurant business per se to compete with our channel. We use name.com to actually make sure that our solution is easy to use. Solve all the problems on our platform before we take it to the channel. Even when we want to do promotions or something, we can actually measure the effect of promotion and take these numbers and share them with our partners so that they know I had a fun when they partner with us. If they do this, what they want to expect and what's the lift that they should expect and all of that. We are very conscious about coming to our channel and having them do things for us. So we try to reduce that lift as much as possible. Got it. All right. So I ask pretty much everyone in the new TLD or in the TLD space this when they come on the show. How many applications do you think there will be this round? Everybody asks you this and it's very hard to tell. I mean, I don't have a crystal ball. But what I think is we're going to see, I assume similar number to the last round. 2000. Yeah, but I think there will be a shifting of demand from where the demand comes from and the type of demand that you get. I think that there are some new names that have been created since the last round that were in our lingo, or a door, or a blockchain and so on. All of these things. So these will be probably highly in demand and highly contested. Roughly 40% of the applications were contested last round. I assume that we probably more than that, this time around, probably more 50 to 55% contested. It's a different kind of dynamic than last round. But overall, I think it would be roughly in the same range. Then probably be more portfolio applicants than last time. Because everybody saw what Donuts did last round and so they probably would like to emulate that. Although I would say the long term is not as rich as it was last round. So I mean, there are different things that are moving parts that you need to consider. But yeah, I think it's probably around the same thing. Well, there'll be fun to see. I've heard everything from Lars Jensen on the program recently will be applying for a lot of domains. He was kind of pegging in that 1250. Most people, I guess it's easy to look at what happened last time. But as you mentioned, this is different. Do you also provide back in for dot brands? Oh, yeah, we do that as well. We think that the brands will be also participating in this round. We think probably more international brands this time around than last time because we didn't have as many if you want in the last round as you know, you'd expect. But still, I think that the brands will participate. I think the generics will participate and then be some Jew key on these as well. So we'll see. And a lot of communities, I think, that we've even seen some pre-announced their kind of community to all these. So great. Well, it'll be a fun year, exciting year. And it'll be fun to watch both what you're doing with DNS ID and as this progresses, as well as when we have revealed a later on this year. And we see what everyone is applied for. I know I'll certainly have lots to write about. Yeah, absolutely. Akram, thanks so much for coming on the show today. Thank you for having me. My guest this week has been Akram Antala. That's a wrap on this week's domain name wire podcast. You can listen to all previous episodes at dnw.com/podcast or on your favorite podcast app. Thanks for listening. [BLANK_AUDIO]