Fintech Insider Podcast by 11:FS
00:25:55 · Kate Moody, Sarah Kocianski, Benjamin Ensor
[00:25:55] S2: He's going to be so cross when he finds out that I've spent most of this episode just berating his shopping. Here we go. [00:26:02] S4: I suppose the way to sum both of those up is under the heading that we've talked about for an awful long time in financial services, which is personalization. So AI can personalize your services to whatever best suit your circumstances. And maybe this is the golden bullet that we've been waiting for for all these years we've talked about the p word. [00:26:20] S6: Bingo. Bingo. [00:26:22] S4: Kate, Ross, do you think the industry is currently paying enough attention to the potential downsides for AI? Too much attention to potential downsides? Which which way do you do you see the industry as as going at the moment, like as a as a whole, as an industry perspective?
00:26:36 · Kate Moody
[00:26:36] S2: I mean, certainly the stuff that I see, maybe I'm in a bubble of caution, but I certainly are seeing lots of you know, the big banks, big Fintechs talking very cautiously about you, yes, understand the opportunity but absolutely wanting to think through how they apply this. Obviously the regulator is playing a very important role in multiple different markets across the world and sort of tempering the speed at which some companies can move. So yeah, I definitely think the downs, I mean maybe the downsides are clear within the industry, whether the downsides have been met, potential downsides have been made clear enough to customers yet, guess is something to, is maybe probably more the point. I'm certainly from what we've seen it sounds like there is ever increasing adoption of AI by individual retailers or business consumers to ask financial questions, receive quasi financial advice. I don't really know if the downsides of those customers are being made clear enough, like whether they are fully understanding that likely we covered on our live after dark show one study that was showing a very significant proportion of those recommendations not being fully correct. Whether that is being surfaced to customers, whether customers aware of that, that if they take action on this conversation, where will the fallback be if something goes wrong if they don't achieve the outcomes they were hoping for? That's probably where the the biggest downside probably still exists. So I don't think it's probably being surfaced clearly enough.
00:28:03 · Sarah Kocianski
[00:28:03] S4: That's interesting. And Ross, how about you? Do you we're in a finance bubble? Mean Perspectively.
00:28:09 · Ross Gallagher
[00:28:09] S5: I think there is, I don't think anybody could doubt that there is, there's almost no regulation whatsoever around AI at the moment more generally. I think again, ultimately where we get to on this in terms of Benjamin's dilemma around, is this gonna do more harm than good? Again, I think comes back to incentives. Right? I think in the early days of DeepMind, they really tried to prioritize AI safety, that being very much at the forefront of how these models got developed. They set up an AI safety coalition to try and bring all of the active players in this space together in a line behind prioritizing AI safety. It never really got off the ground because of course the incentives were around, well, who's gonna win this AI race? We've all heard Donald Trump say whoever wins AI wins, And now they're talking about these self governing coalitions. I mean, think what you will about the idea of self governing. But I think, look, the more nuanced view on that or maybe the more practical view is you know, if if if people are gonna be using some of these general purpose frontier models for how they manage their personal finances, there's so many questions I think that needs to be answered. And I think one of the the most important questions is, know, whose behalf is the agent acting on? You know, whose outcomes are the agents prioritizing? Are they prioritizing delivering good outcomes for the consumers? Are they prioritizing delivering the best possible outcome for the AI companies? You know, there's so many questions here I think that need to be addressed before we can start to get under the hood of does this get us to a place that's better off or worse off? But boy, does it have the potential.
00:29:50 · Benjamin Ensor
[00:29:50] S6: I think the fundamental challenge here is twenty, thirty years ago, financial advice and financial guidance was all quite tightly regulated in every country. And the only other people who really ever talked about financial decisions were a small number of financial journalists, right? Magazines, newspapers, and so on. Now, everyone is a publisher, right? Everyone can publish whatever they like on the Internet and that's wonderful. And there's all sorts of weird and wonderful things that everyone publishes. However, it creates this path for scammers and all sorts of bad actors. And the problem is there's no regulation of that. So the regulators have taken this sort of relatively straightforward view of, okay, we need to regulate financial services firms. Well, unfortunately, no. That's not good enough. And so there's a massive issue, not so much for regulators, but actually for legislatures and political leaders of how do we regulate AI and media more widely because is it okay for individuals to be going out and giving advice to people on, you know, whatever pyramid investment scheme or whatever they've come up with. It would be very sophisticated, very slick advice, but it completely unregulated and outside the remit of the regulator because it's not a financial services firm doing it. And so I think that's actually the biggest problem is the legislators need to really grapple with this of how you're gonna tackle this. Oh, and just to make it more complicated, let's throw an extraterritoriality as you just alluded to. If these AI models are coming out of China or The United States and you're in Canada or Ghana or Australia or wherever, well, can your regulator really do very much to prevent the actions being done by an AI company in a completely different country on the other side of the world. So that's one heck of a big issue. And that's why I see this as having potential for huge harm because the regulators don't necessarily have the tools they need to prevent harm to people in their countries from bad actors in other countries.
00:31:50 · Sarah Kocianski, Kate Moody
[00:31:50] S4: It's certainly a very big topic and not one that's, you know, unique to any particular country, region, or otherwise, as you suggest. So I'm sure that we're gonna be hearing a lot more about that, over the coming months and years. [00:32:04] S2: You can just wave your magic wand, though. Right? You can just fix it right now. [00:32:07] S4: Oh my. I tell you what. There's other things that would be higher on my list first. Right. We've heard about three very different wishes today, but making them come true, as we've discussed, is not gonna be quite so simple. So let's talk a little bit about what's actually standing in the way. So we've alluded to to some of these areas already, but we've talked about better incentives for banks. We've talked about making money easier to talk about, and we've talked about ensuring AI benefits customers. So what do you think is the biggest thing getting in the way? So we've we've touched on lots of different areas and potential, you know, barriers for these things happening. But if I asked you for each of your wishes, what is the biggest of the barriers? What would you say? I'll come back to you first, Kate, because you just started.