Welcome to the Connected Podcast with Charlie Taylor.
Hello there. Great to be with you after being away for the last few weeks on holiday. Back in time for the cold to hit us here in Dublin. I'm joined this week on the podcast by Vish Gain, one of our reporters in the newsroom. Hello Vish, great to have you with us.
Great to have you back, Charlie.
Brilliant. [laughter] Now later I'll be talking to Sarah Friar, the Irish-born chief financial officer of OpenAI, one of the hottest, if not the hottest, tech firms on the planet right now. But first, it's a big week for Irish tech as Dublin plays host to the International AI Summit, with leading figures from the sector in town to talk about what is the hottest technology bar none right now. Now among those speaking at the summit are going to be Demis Hassabis, chairman and founder of Google DeepMind, Sarah Friar, who we've already touched on, and Brad Smith, president of Microsoft. Now, Vish, there's plenty of news coming out from the summit already ahead of it happening on Wednesday at the RDS. We have Google, they've just announced that they are going to be investing 5 million to help advance AI-focused weather forecasting in Ireland. And OpenAI itself has announced 200 additional jobs. This is in addition to the 250 jobs that were announced in July. And Tines, the lovely Irish tech unicorn that we love so much, has also announced a hundred jobs, with most of those going for entry-level staff as well, which is pleasing to see. What do you think? It looks as though it's a big event. Will it do well for Ireland in terms of putting us on the map globally?
I think so. I mean, it's a testament to the fact that Ireland is already on the map globally in terms of AI. That, you know, 200 jobs from OpenAI coming after a couple of months after 250 jobs by OpenAI in a country where they only have 120 at the minute is a huge investment. That's a sign that this is where they hope to grow. It's not just about tax anymore. It's about talent and it's about the fact that this is seen as the place to be in Europe when it comes to AI. So that's very refreshing to see. It's not just OpenAI. There's homegrown Irish companies such as Tines, as you mentioned, that's also doing that on both sides of the US. So there's Stripe that's making jobs in the US and coming to Ireland and making jobs, but there's Tines that's going to the US to do jobs as well. So they've announced loads of jobs in the US.
They announced 100 jobs in Boston, I think it was earlier this year as well, headquarters. And obviously we can't ignore Anthropic, which is also growing very fast in Ireland as well. So I mean, it's looking good in terms of that. Now you also have some numbers, we had some numbers out this week in terms of how Irish tech is doing in relation to getting VC funded.
Yeah, it's huge. So the numbers, there's PitchBook data that's come out that's based on a report that was commissioned by Enterprise Ireland. And it shows that VC investment into Irish AI firms has reached 950 million since 2025. So in the past 19 months, which is huge, 950 million, that's nearly a billion for just Irish AI firms. That's a specific niche within the Irish tech sector. So that's a lot of money. And this includes some big names. Tines itself, 120.7 million in a Series C round last year.
That was what tipped them into being a tech unicorn, of course.
Yeah. And they were also in Sifted's top 100 or top 250, I believe, including some other names as well. Seven Irish names in total, which was great to see in the top 250 of the fastest-growing startups in Europe. But there's also Camana, 49 million. There was Fenoa, the Irish startup, that's 96 million. And in total, there were 182 deals in the past 19 months that accounted for the 950 million in funding. So last year, if you break that up between last year and this year, it was 691.3 million in 2025, which was record-breaking in terms of Irish AI investment. Now, it's not surprising that it's record-breaking because, of course, it's growing now. It's not like AI has been around for 20 years. So every year will potentially be a record-breaking year, just like Nvidia's results every quarter.
And still not enough for investors, as we often...
I'm disappointed with 950 million myself. Sorry.
Why couldn't they at least push it over the billion mark? Yeah.
Exactly. Yeah. And 200 jobs by OpenAI. Not enough.
Very good.
And it includes not just AI companies that build AI, but AI-adjacent companies as well. So companies that are part of the AI ecosystem. So it could be a medtech company, but a company that deals in AI as well.
Very good. And so, and now to be honest, I mean, is any sector not an AI company?
Those that are focused on AI are doing very well. I mean, there's been big investments in companies that are focused on AI. Obviously we touched on the likes of Tines and stuff, but LearnUpon this week also announcing 50 million in new financing as well, which is from K1 Capital. It's a few years since they raised funding before, but that's all for them pivoting towards AI all the way for its learning management solutions. So, you know, anything that's with AI on top seems to be doing very well. I suppose, you know, we'll go for our break now. We are going to... I'll be back in a second to talk with Sarah Friar, the Irish-born chief financial officer of OpenAI. And for anyone that isn't at the AI summit this week, we will be covering it in depth in the Business Post. All the news, all the features, all the interviews. Both myself and Vish are going to be there, aren't we, covering all the bases. So be sure to look out for that now. [music] But straight after this, I'm chatting with Sarah.
Thanks ever so much for taking the time out to chat. As you know, I know you've got a crazy busy schedule. So I suppose to start off with, I mean, let's cut to the chase in terms of Donald Trump. He's just said, you know, we should rename AI as superintelligence. Does that mean there's going to be a rebrand of the company coming into life?
Well, I think overall, if you go back, OpenAI, our mission is AGI, so artificial general intelligence for the benefit of humanity. And I think we just never leave that behind. That's the most important thing, is always at the end of the day showing up to kind of help people in their day-to-day lives, whether it's fixing their car troubles like we just talked about, maybe something they're doing on the medical side, and then all the companies that we get to work with. It's incredible.
That's the focus. Yeah, that makes sense. And come here and tell us, you arrive in Dublin. I know you're a frequent enough visitor given this is your home country, but also, you know, in terms of, you know, you've come this time bearing gifts, 200 new jobs. Obviously we're delighted to hear that. That's on top of 250 jobs announced earlier this year. What is the thinking behind, you know, ramping up particularly in Dublin?
Sure. So it's actually an easy decision to make overall. Like number one, we find, you know, ChatGPT is used around the world, 1.2 billion people all over the globe. Ireland itself is a great user of the product. I think we have 2.5 million people use it every week, which I think is almost like half the internet population and growing at a massive pace. So, number one, people use the product, they like it. It's always good to be in a place where your employees, you know, when you put your employees in place, that the product's being used. I think secondarily, we've always viewed Ireland as a tech hub. You've seen it in every single tech cycle. I couldn't be prouder as a Tramore woman to see Ireland always rise to the next generation. What we see here is a population that's highly educated, is tech-forward, is very open. We like the brain. It's no longer a drain. It's a brain suck back. We already had some, you know, we already had an office footprint here that was doing well for us. We're at about almost close to 200 people. We did make the announcement in July for 250 more, and then with this recent 200, it's really trying to say these are technical jobs as well. So what do I mean by that? In OpenAI, we're seeing this rise of something called forward-deployed engineers. What are those FDEs doing? They're really helping companies begin the transformation journey with AI. So it's not, you know, it's hard to show up and be like, here's a model, and people are like, I don't even know what a model is. What does that do for me? Instead, you want to be talking to, you know, an airline about how are you going to do faster scheduling, a bank, how are you going to do better customer service. So those FDEs help do that transformation. We already have something going with the University of Limerick, so it's an easy place to get started, but we'll do it with all of the universities. And we couldn't be more excited. More people, and if you're listening and you want to be an FDE, give us a call.
And I wondered if there's a possibility of more entry-level jobs coming along somewhere down the line, just because, I mean, obviously, as you know, Ireland and everywhere else is trying to get up on AI, and, you know, in particular, people are looking for roots into the sector.
Totally. Well, the FDE program, you know, today we're looking to hire those 200 people. So we are looking for people that are through university, probably have some experience to date. But that's why I mentioned University of Limerick and the university system, because if we can get it into university courses and it can be technical colleges, doesn't have to only be at one level of a tertiary education. We will start to see people get trained early. They can be our interns. I love a good intern. Hopefully they maybe have an internship or two with us and ultimately hit the workforce. Like to me, mass success is we are sitting here having this conversation in two, three, four years from now, and I'm telling you our entry class at OpenAI, 50% of them went through this program. Like that will be not just that we're hiring people in the Irish population, but we're actually training the up-and-comers because to your point, AI is going to bring about a whole new transformation of job types too.
Excellent. And tell us, you know, obviously you're here for the International AI Summit. You know, I often ask you anytime I meet you, it's kind of like, how do you think Ireland is doing in terms of AI? So here I go with that question again. How are we looking? Have there been improvements since, you know, I mean, you're obviously keeping an eye on things from where you are in the US.
Yeah. So if you look at overall, we just have a new economic research piece coming out. If you look at countries that are on the leading edge, it's interesting. It's a lot of Nordic countries, a lot of the Balkans, so like Estonia, Lithuania, but also Ireland is right up there. And again, it goes back to, I think, good investment at a government level into staying on the forefront of where technology is going. One of the two of the areas we like a lot here in Ireland are chip development. That's a place where models are getting smarter and smarter. And in the US, we're starting to work, you know, we've built our own proprietary chip, in fact, which taught us a lot. But now we're working with a lot of the various chip companies like Nvidia or AMD, for example, here in Ireland, as you know. I think that's kind of blown people's minds that there's so much chip expertise, particularly in a world where people are trying to make sure for geopolitical reasons that they have that expertise sitting here in Europe. So that's one area where we're excited to benefit working alongside. And then the second area is cyber. So, like, if you ask me what keeps me awake at night, which might be one of your questions, I am worried about how cyber-strong these models have become. I know what our own models are capable of. At the same time, Chinese models have become very, very strong on the cyber front. So a lot of the offense already has tools in their hands, and we need to make sure the defenders have better tools in their hands. And cyber is a place that Ireland as a country and as a region in Europe has put a lot of investment into as well. So those were two of the things that really drew us to Ireland to put this FDE program in place. So we see good alignment from government, good alignment from the private sector, good alignment coming from just the education sector. And at large, Ireland also acts as a really nice footprint into this large European market, right? In and of itself, Ireland is a smaller market. So this is really about how do we then go talk to companies throughout Europe in order to be able to help them through their AI transformations.
You mentioned China there, and obviously the Trump administration has gone big on the thing of like, we need to keep moving forward in terms of progression on AI, partly because otherwise China will win the race. Is that overblown, do you think? Is that really where the race is in terms of AI?
I think it is definitely not overblown. Absolutely. If you look at the frontier labs today, they're largely sitting in the US and then China. I think everyone views it as a strong national security asset to have frontier models. And to that end, you know, we agree at OpenAI, that's why we are doing our best to keep investing to stay frontier. And on the cyber front, why we're trying to help our customers. Like, I don't like selling fear. That is a really horrible way to sell, and it's not usually a long-term way to build partnership with people. You need to sell solutions. So actually, we just had a conversation with a customer sitting in this chair that you are in right now five minutes ago, talking about how we actually practiced a little on ourselves first. So with Astra, our frontier model, we turned Astra on our own infrastructure about six weeks ago. We looked for all the holy code, I call it, all the places where there's vulnerabilities. From there, we built, think about it as almost a log, a table of here are the vulnerabilities, now we need to go fix them. So Astra actually helped push it out to the right owners throughout the company, track that the changes were happening, that people were building patches, helping QA, so quality assure those patches, so that when you roll them out, you don't break [clears throat] your underlying IT infrastructure, and then say, okay, all clear, check.
We've had incidents recently, like the Hugging Face incident, you know, and obviously a lot of other companies have come out now and said about instances that they've had, you know, AI models go rogue or however you want to put it. I, you know, obviously that has led to, you know, consideration being given of should we slow down in terms of frontier models. You know, what Sam Altman has joined with Dario Amodei, Elon Musk, everyone seemed to be singing off the same sheet, albeit very briefly. Do you see more of that sort of, you know, partnering, I suppose, in terms of, you know, addressing concerns that are happening?
So at the ecosystem level, we absolutely, I think, need to work as a single ecosystem for the benefit, and going back to our mission statement, right? The benefit of humanity, but I would call it for the good of all the companies and consumers that we work with. Like, I started my career in Infosec, in cybersecurity, interestingly, right at that moment where we'd connected the internet and suddenly we had this thing called viruses and no one really knew what they were. And then we had this big moment called Code Red, for those of your listeners that might know about it, where suddenly it wasn't what a virus was. It was in my day-to-day life, if I called 999, the ambulance didn't show up. If I tried to take money out of the ATM, my card didn't work. If I tried to get on a plane, the booking system was down. And that's when you went from kind of [clears throat] tech speak to how's this really going to feel to me in real life. So that was at the beginning of the internet. We have solved a lot of the virus issues, right? Most people would not buy a computer today that didn't have antivirus technology deployed on it. If you kind of go through then the two decades since that happened, we've seen antivirus intrusions happen and get stopped. And I think that this is just another era in some ways of the climb. It's not just going faster. And I think the intelligence behind it is greater. So for sure, Hugging Face for me is one of those moments where, you know, never waste a good crisis. It's absolutely a call to action for the industry. It's what I just described on let's patch all those holes in code. But I do think that we're moving into a new era where we're going to think about what can this technology do with agents? What sort of security do we need to make sure that agents are working on behalf of humans? And I think that's something that OpenAI takes very seriously, that there's a human in the loop, that this technology will move forward with humanity side by side. But we do want to work with our ecosystem too. In fact, we put out a whole blog post, Hugging Face, with, I think, over a thousand companies included, and that included Anthropic, it included big companies like Microsoft and Amazon, but all the way through the whole security ecosystem as well.
