Q+A: Telstra’s AI strategy explained

0:00Welcome to Fear and Greed Q&A, where we ask and answer questions about business investing, economics, politics, and more. I'm Sean Aelma. Artificial intelligence is quickly becoming one of the biggest strategic investments for major companies, and Telstra is performing a balancing act, expanding its use of AI to improve efficiency and customer service while not over capitalising on the technology. Michael Ackland is Telstra's chief financial officer. He's also just won the Investor Engagement Excellence Award at the CFO Awards. Michael, welcome to Fear and Greed Q&A. Thanks, Sean. First, I'm just on the award, the Investor Engagement Award. How the hell do you engage with what more than a million investors? That is quite the job you've got. Yeah, absolutely. And luckily, we have a whole lot of analysts and institutional investors that we can engage with deeply, and they amplify our communication around what we're doing.

1:00But it's a big job, and it's all about alignment, Sean, and getting making sure that our message is aligned with Australia's year, and we do what we say we're going to do. So it's like many other parts of life. One thing about Telstra, there are a lot of your shareholders who want a dividend, expect a dividend, and in the last results, last month they got an increase, dividend, they got a share buyback. I don't know whether they're happy with share buybacks or whether they prefer dividends, whatever. In real terms, how do you balance all that, because obviously you've got institutional shareholders, but you've got a lot of very small shareholders, how do you balance those two things? Yeah, well, I think we've had a pretty clear investment narrative for some time, and you rightly point out the dividend. Many people own us for that dividend, and they own us as a defensive stock, and a stock that's going to be a reliable cash flow generator, and one that they can trust, and that's really important. And we think about that when we think about our strategy, we're in it for the long term. I like to say we're here for the shareholders that are going to be the holders, not those

2:05that are in for a short time, but a good time. And I think that's really important with a stock and a company like Telstra without history and our assets, that we really align those things around being there for long term investors and long term holders of our stock. I mean, and I'll get into AI in a moment, just one final thing though, it's sort of on that corporate strategy. You kind of sit between the market and the company being a CFO. When investors are looking at Telstra right now, and you've had certainly in the mobile space you've done really well, other spaces a bit more difficult, what are investors talking to you about? Yeah, well, it is 16, and we will get to AI, as you said. They talked, they asked a lot of questions about AI. Yeah, absolutely. But I think they're also always trying to understand how our physical infrastructure plays into our business. So that physical connectivity infrastructure that spreads out all over the country, all those ducks and pits and pipes and cables and towers and all of that that just sort of

3:05makes that magic happen that we all experience when we pick up our phone. They really sort of tease that out. How big is that, how big is that moat, how could that get disrupted by different players? And I think they really make a bet on that asset base that we have, and just how important that is for the economy and for communities and for all of us in the way that we connect every day. Okay. Let's talk AI. I am sure that there are a lot of boards and senior management teams around the country who have invested plenty in AI, and they've probably said to management, hey, we've got this great AI now. Make it work. Right? Now, there's going to be a disconnect there at times because the world of AI, and I think we're talking to you because you have really distinguished between, I suppose, the reality of what you can do and that vague promise of AI changing the world. Yeah. And it's like a bet you make on any technology. You know, first of all, you need to make sure that the first principle of what you're doing

4:08is you're aligned to your strategy and you're making sure that that technology, the technology isn't the strategy. Your strategy is the strategy and the technology supports it. And I think, you know, that's been real important. One of the things we think about with AI, we are making a bet on it. We want to be an AI first company. We want to make sure that we're brilliant utilizing AI. And so there's probably two things that we think about. One is AI companies are making a mountain of investment. I think the estimate is over 700 billion dollars of cat-backs this year across those AI companies. That investment is going to look for a return. And we experienced this with cloud. So being really disciplined on our architecture, making sure that we don't lock ourselves into single vendors, that we set ourselves up so that we can swap between different vendors, that we understand the costs of the tokens and the storage and the cloud compute. So making sure that we've been really thinking about that deeply and we engage all over the world with experts to make sure that we're really understanding where people are thinking there. The second one is just around that balance between, you want to sort of send out and get

5:12people to innovate. So we did that by giving lots and lots of people access to co-pilot and training our people with our AI academy. And you sort of let a thousand fires burn for a while, but then you have to start to focus down on the areas where you really can see the value. And I mean, we're looking at, like everyone, customer engagement, how you deal with customers, AI is going to be a big opportunity there. And we're seeing real benefits in co-sign our handle times in people using generative AI chatbots and more stuff happening on digital. But we also look in our network operations, you know, and how we run the network. Can we make faster decisions in the network, respond to alarms faster, optimize our network in different ways, as well as just in that hole. And we've all heard this recently around the way it's fundamentally changing the way we develop software. We spend about a billion dollars a year internally developing software, spend on IT, and we're seeing significant changes in that space as well as people use it to do coding and speed up the way they do coding. And we've seen out with a lot of software companies as well.

6:13So disciplined on the infrastructure we set up to exploit AI, we're not an AI company, we're going to be an AI first company using it, and then disciplined, like developing capabilities in our people, but then being focused on where we think we can get the benefits. So I get what you're saying, let's talk about the first, I mean, the chatbot example to begin with, and we all go to chat GP, well, maybe don't, but most of us go to chat GPT now and think, I've got a headache, what's it mean? It's kind of a similar sort of thing. I have a problem with a mobile or a fixed land line, and I go to chat GPT, and it helps it, or I go to Telstra using one of the larger language models, and it helps. How good is that right now, because I'd say five years ago or two years ago, if I rang Telstra and ended up in EQ, I'd be frustrated, they'd take a long time, like how much difference has that made? Yeah, so there's two things. One is we've got a chatbot that supports our call center agents when they're talking to you, and we've seen, you know, probably, you know, 20 to 30% reduction

7:16in answer times on, on life for like calls, as the agents are better at answering the questions, and they're more accurate in answering the questions. Yeah, but we had to clean up a lot of our data to make that chatbot work, and then the other one that we only launched recently was our first generative AI chatbot for customers, that's at the moment through the website, it'll be launched through the app soon, and that compares to the old chatbots that we experienced, which were more machine learning, sort of linear thinking, which weren't great for many of us if you got stuck in the wrong doom loop, you couldn't get out and trouble. So what we've seen with the generative AI chatbot with customers is we've seen the containment rates, so the number of people who resolve their issue in there and say to satisfy triple. So that's really exciting. We've gone from probably, you know, 15% of people who were able to satisfy themselves within the chatbot, and I'll have to go to a person to now about 45% with generative AI, and I think that'll get better about, we're all experiencing that with when we're using chatGPT or Claude or Microsoft Copilot,

8:21we're seeing that it's getting better and better, and we're also seeing that in our ability to apply it with customers. At the enterprise level then, how do you know where AI will work for you, and where it can add value? Because there are a lot of areas where it won't work for you, I'm guessing. Yeah, absolutely. And this is where it's so important to, the most important capability with AI is your own people who know your processes, who know your business, and so that's why the first step we took on AI, well, the first step was, we got very focused on cleaning up our data, we thought that was really important in simplifying our environment. But training our stuff, and we've partnered with Accenture, which has been well publicized with a joint venture with 12 or 13 hundred people in that joint venture, which are our employees and their employees who are going after this. But training across our sort of 20,000 workforce is training the AI, giving them access to a corporate level co-pilot so they can try it. And I

9:22really believe that it's our people who are going to tell us whether or not AI is going to work for those processes or not. And it's only by getting more and more of your staff engaged, learning, experimenting both privately, but also in the workplace that I think they're going to tell us what works and what doesn't. You've talked about over capitalising on AI as well. How do you know whether you're over capitalising on AI? Well, like most have over capitalising experiences, you know afterwards. Yeah. Hindsight. Hindsight is a wonderful thing. I mean, it's absolutely the crunch issue because it's a classic, you don't want to move too slow, you don't want to move too fast. And I think the thing that we're focused on, I'll go back to that, is being really disciplined on the architecture we set up. So one of the things we've done within our extension JV is to start to look at what we call an AI control plane, which is giving us greater insight into what's the right LLM to use for what outcome to manage and track token costs

10:29which could get out of control to be able to plug in and plug out of things when they're either working or not working so that we can control that cost curve of capitalisation as we go for it. But I think the point is you won't know and the real trick is to keep yourself as flexible as possible so that you can react as you learn more. Yeah. And so the point there is having not being tied to one supplier for the next five years actually moving between as necessary. Yes, absolutely. Keep that commercial tension, keep those optionality. At one final question, it always comes back to staff and Telstra being such an iconic company. How will you manage that? There will be job losses, there has to be, but there'll also be job gains on the other hand in that area. How do you manage that as a company? Yeah, absolutely. And what we're focused on is training staff, is giving staff that capabilities so that they're more capable in an AI world and an AI era. And I think that's all you can do is help prepare people to be as capable as responding to what will be a new environment as you can.

11:34Obviously, any decision you make where staffing levels move up or down is something that is taken incredibly seriously. And we've got, as I said, as a company, brilliant physical assets and unique physical assets that are core to our competitive advantage. But we've also got tens of thousands of really passionate staff that are really core to our competitive advantage as well. And we recognise that history of what Telstra has been for people and what it will continue to be for our staff. And so investing in those capabilities, I think, is sort of the best lever that companies can take to make sure that one, as I said before, they're better at using the AI, because if your people know how to use it, they're going to be better at identifying the opportunities. But second, setting them up in the best way for the future. Michael, thanks for talking to Fear and Greed. Absolute pleasure, thanks, Sean. That was Telstra CFO, Michael Ackland, I'm Sean Elmer, and this is Fear and Greed, Q&A.