Life Insurance: Consumer Duty Teardown #4

Episode transcript

Amelia: Hi, I'm Amelia.

Russ: I'm Russ.

Pat: And I'm Pat.

Amelia: And this is Fin the Week. How is everyone?

Pat: Good.

Russ: Yeah, good thanks. I think I had my best LinkedIn post impression rate this week. It's quite good.

Amelia: What was it about?

Russ: It wasn't about website design or app design, it was UX — but UX for baby clothes.

Amelia: I saw this actually.

Russ: It's had over 33,000 impressions so far, about 250 engagements, and every time I log in someone else has liked it from around the world.

Amelia: That's amazing. For anyone who hasn't seen it or doesn't follow you on LinkedIn, what is it about?

Russ: Okay, maybe we should change the title of this episode actually, but — it's about baby grows. You get two different coloured poppers for the male and female poppers. They're normally orange or brass, and if you start with those first, all the rest of the poppers snap into place. It's quite confusing, especially if you're doing it at 2 or 3am like I had to last night — it's difficult to line them all up, especially when the baby's wriggling around. So yeah, it's touched a whole audience of parents, and also people who had no idea these handy little features existed. I wasn't expecting it to go viral, but it did. There you go.

Pat: Nothing like tugging on people's heartstrings to generate impressions.

Russ: It's a great design feature, and it's actually given me an idea for more LinkedIn content — just looking at really simple, good design features that help people in their lives. Not just highlighting them as inspiring, so people can bring that into their work, but celebrating them really. So I'll try and find more.

Amelia: Everyone connect with Russ on LinkedIn for more of this content. But no, it obviously struck a chord with a lot of people.

Russ: Yeah.

Amelia: So it's officially the anniversary today — three years since Consumer Duty came in for open products. This is the final Teardown episode of anniversary month. Russ, for anyone who hasn't caught the last few episodes, do you want to briefly explain what these Teardown sessions actually are?

Russ: Yeah, we've been looking at different segments in the finance sector. We started a while back with robo-advisors for investment, then mortgage brokers, and bank switching for the first one. What we're doing is looking at the journey of someone learning about these websites, choosing a product or finding advice, and perhaps contacting the firm — and seeing whether that journey aligns with Consumer Duty. We use our tool, UX Duty, which automates the scanning of these websites and gives us back a set of recommendations, and then we discuss them on the podcast.

Amelia: So that's what we've done so far, and we've saved arguably the most regulator-charged sector for the last week: protection insurance, specifically life cover. Why are we looking at this?

Russ: The FCA recently announced the areas where they have active open investigations, so we thought we'd look at the insurance sector, because seven out of the eleven open FCA investigations are in insurance. It felt like a good way to end, because — although they haven't issued any fines yet — it's clearly a big focus of theirs.

Amelia: So let's jump in. What types of firms have we looked at this week?

Russ: In previous episodes we've looked at different-sized firms, but we took a slightly different approach for this one. We looked at firms that offer life insurance in different ways, to give a broad view of the landscape and see if there were common trends. Again, we're not naming the firms — similar to how the FCA communicates its investigations. So we have "Firm A", a direct insurer you'd go to directly for cover. "Firm B", a comparison site or aggregator — you land on the site and it compares different direct insurers to find you the best deal. And "Firm C", an independent financial adviser, or broker. So three different ways of finding an insurance product across three different firms.

Amelia: What process did you take them through?

Russ: We picked the journey a user would actually go on: land on the site, look for a product, try to understand what they're buying, then start the application process. There are normally six or seven URLs in that journey, and our tool, UX Duty, scans those URLs. We then validate the findings and pull out common themes — looking at each URL and analysing it against the full Consumer Duty outcomes: products and services, price and value, consumer understanding, and consumer support.

Amelia: So that's what you've done. As always, let's start with the overview of the headline findings.

Russ: 109 findings across the three firms, and the common trend is that around half of those are accessibility findings — again. All of them came back as high risk, and it was a fairly equal split in terms of findings per firm. Firm A — interestingly, although it's not the largest firm in this round, it had the most findings this time. Previously it was the biggest firm that had the most. This time it's the direct insurer with the most findings, followed by the broker, then the comparison site. I think everything comes back as high risk because if there's anything that could be perceived as serious, it's a compliance risk, and that's naturally sensitive — you wouldn't want to label something as medium if there's a potential breach. It's a fairly common approach across compliance teams, as we've discussed before.

Amelia: Is there anything that surprised you?

Russ: Because we've done this a few times now, the accessibility findings weren't a huge shock, and I've got a summary at the end covering accessibility across all the firms. There are some themes we've seen from previous teardowns too. Nothing hugely surprising, but I'll go into detail.

Amelia: Should we get into it then?

Pat: Sure — I was going to say there's one finding that's surprising in that it's kind of sadly unsurprising. We'll come on to that in a minute.

Russ: Yeah, I think I know which one that might be.

Amelia: Little tease there. So should we dive in? What's the first finding you wanted to discuss?

Russ: You land on these websites and see a big headline: "get your life insurance for £3 a month," "£4 a month," "£6 a month." It feels designed purely to draw you into the process, and from a Consumer Duty perspective it's not clear who's actually eligible for that price. It's a common trend — for example, Firm B has a headline of £5 a month for cover, with the dreaded asterisk. We've seen this on bank switching, on mortgages — where's the explanation for the asterisk? You have to find it on the page, normally in the footer. You can search for it if you're savvy — Command+F — but it's usually hidden away, and hidden small.

What's confusing comparing the three sites is that they each construct that headline number differently. One firm has a low monthly rate with an asterisk explaining it's based on a 25-year-old non-smoker insured for £100,000. Another has a headline rate of £5 a month, with a note that it applies to 50% of their 30-year-old audience, and doesn't include critical illness cover. So you get this big loud headline, but no real explanation of whether it aligns with your situation — it's almost meaningless. These websites lead with it. It's psychological: you land, you perceive it as affordable, you get drawn into the process, and then you realise it could be ten times that amount. From a Consumer Duty perspective, they're not specifying the target market at a granular level at the first point of the product life cycle. That was the first thing I noticed, and it's definitely a common theme — big headline costs with no transparency about whether they're for you, or about hidden terms.

Pat: So what happens is: you go to the website, see the headline, and the vast majority of people believe it. They make an enquiry, a salesperson calls them, gets them into the system, realises the person is 38 and has, say, type 2 diabetes or has been in hospital for this or that, and suddenly it's £98 a month — which is much more representative. £6 a month for £200k of cover is targeting a tiny proportion of the population. Most people aren't 25.

Amelia: Do you think it actually matters? Once someone's in the system and has started the process, don't they usually think, "well, it was a bit too good to be true"? Do you think people are being drawn in by the headline, and is anyone actually reading the small print?

Pat: I guess it's about finding the best deal. If a user has a long phone call with a salesperson, are they going to want to repeat that five times with five different providers? Probably not — they'll likely just go with whoever they started with, and as a result they won't get the best deal.

Russ: And is it in line with Consumer Duty, where price and value is one of the outcomes? You could argue the firm wants to get business, so they put the most attractive rate up front to draw people in, and people figure it out as they go — but firms have a duty to the consumer to have fair and transparent pricing end to end. If it's misleading, it isn't in the spirit of Consumer Duty. I don't think the FCA would intervene on it directly, because people do figure out the real deal eventually, unless someone rings up and says "I saw your £3 a month, sign me up for whatever" and then gets charged £100 a month — that would be a bigger issue, but there are processes in place to stop that.

That was one of the first findings. Also, quite often, the small print is under the text size we'd expect for accessibility. We've talked about 16 pixels before — last week Gov.uk moved their design system up from 14 to 16, and the British Dyslexia Association recommends never going below 16. We saw 9 pixels last time; this time, for Firm B especially, where the headline had "no critical illness cover included" in small text, it was 13 pixels. Not as small as we've seen before, but for something this important — especially for vulnerable customers — sticking to 16 pixels really matters from a UX perspective.

Amelia: Shall we move on to the next finding — this one's about consent?

Russ: Yeah, I think Pat's best placed to cover this one. If anyone's tuning in and noticed, Paul's not around this week, so we're a little all over the place without him.

Pat: Yeah, this is the one that's sadly unsurprising, because it's a flagrant breach of GDPR, and you see it everywhere. We talked about it last week too: Firm A's homepage has a consent or marketing checkbox pre-ticked on page load, meaning the customer hasn't opted in — they have to opt out, which is explicitly against GDPR. The reason companies do this is that if you ask people to opt out, maybe 80% won't bother, and your mailing list grows nicely. A mailing list is a powerful marketing tool with a good conversion rate. But if you ask people to opt in, maybe 5% will actually check the box, so your list barely grows. The incentive to pre-tick that marketing checkbox is really compelling — even though it's illegal — and you still see it on regulated firms' websites.

Amelia: Are there repercussions for this? Is it just that big firms know they'll get away with it?

Pat: It's largely unenforced. There was a case in Guernsey a couple of years ago under Guernsey data protection law, where a customer was added to a mailing list without permission and started getting marketing emails. The local data protection office took the company to court and issued a £10,000 fine — but that's the exception. Probably 20–30%, maybe more, of marketing checkboxes online aren't implemented properly, and enforcing that across the board is basically impossible. I think regulators wait for customers to complain, then reach out to the provider and say "you need to fix this" — and that's usually the end of it. I don't think people even realise it's illegal to configure checkboxes that way.

I'd be interested to know, Russ, whether Firm B or C had confusing wording too — double negatives, "uncheck this box if you do NOT want to be on the mailing list," so people accidentally opt themselves in. That's illegal as well, because consent needs to be clear. And Firm C bundles the privacy policy, T&Cs, and "about us" into a single checkbox, so the customer can't meaningfully consent to any one thing. That's pretty serious — with consent, you need to understand specifically what you're agreeing to. It should be written in plain terms: "I consent to [Firm] using my details to send me promotional emails related to campaigns we're running, in line with our privacy policy" — and then link to the policy. The core purpose you're opting into should be in that message.

Amelia: It feels very sneaky. I'd love to know how these conversations go internally — "how do we word this so it doesn't sound misleading... but is it misleading enough to work?"

Pat: Most of the time there's in-house counsel or a lawyer saying "you can't have that pre-ticked," or an agency recommending it be unchecked by default — and a marketing director on the other side saying "I see it everywhere, we can get away with it, and it's really powerful." There's tension between those two factions in these firms, and in some, legal wins; in others, marketing wins. That's probably how it pans out.

Russ: What's the solution here? Cognitively it's very confusing, and it's not a one-off — almost every site uses double negatives to try to trick you into signing up. What can users actually do to make the web a bit easier? Complain, I guess, going by that case — but people won't complain because it feels too minor, and nobody wants to go through a legal process over something that feels small. I wonder if AI could help — our scanning tool could call out every instance, produce a report, and send it to the FCA or relevant authority to work through. Though that would be a big job unless you sent an AI agent to do it.

Pat: I think we'll see browsers that lead on privacy — Firefox, Brave — start integrating LLM-based processing into browsing, so it can actively navigate you out of confusing checkboxes. I also think we'll see a rise in AI agents filling out forms on your behalf. You can already do this — Claude has computer use, ChatGPT has computer use. I haven't used it loads because it's slow, but I know people who routinely ask Claude to fill out forms for them, and even though it's slow, they're happy to let it run in the background. An AI agent will do a much better job of spotting double negatives and confusing wording, and opting you out of everything you actually want to be opted out of. Over time this tactic will become less useful for collecting mailing lists.

I also think — probably a separate episode's worth of discussion — that email marketing itself is going to get compromised. I've seen a huge increase in marketing emails hitting my inbox that aren't caught by spam filters, because there are tools that use AI to scrape data and mass-email people who never opted in. I think we'll see a rise in AI scanning inboxes for unsolicited marketing and just deleting it. Every morning I've got a wall of spam — my Gmail spam folder has around a hundred messages, and another twenty in my main inbox that Gmail didn't catch, and Gmail's spam filter is best-in-class. Two or three months ago it wasn't like that, and it's only going to get worse. I'm actively looking at running an LLM across my inbox to filter this noise out. If you haven't opted into something, it's going to get harder and harder to reach that person.

Russ: One of the challenges is that both the attack and the defence are getting more sophisticated — we'll probably end up with forms designed by AI specifically to get around the AI that's trying to stop people signing up. It's going to be difficult for this to go away for that reason. Shall I move on to the next point while I'm in flow, since it ties in?

Amelia: Yeah, go for it.

Russ: I have a theory — it'd be interesting to see the conversion rates on these sites, especially the comparison site. I suspect collecting personal data is high up their agenda, given the pre-ticked consent boxes and small-print terms. What's the actual conversion rate on signing up for a life insurance product versus the value of the data collected along the way? Is it being sold to a third party? Used for marketing? Another revenue stream? Data is powerful in this world, and if they're making sign-up complicated while hiding the terms about what they do with it... Last week we also found a complaints form that required emailing them just to opt out of the marketing list after submitting it.

So the third theme is: it feels like every firm is collecting personal data before telling the customer whether the product is even right for them — which comes back to eligibility. Consumer Duty is clear that firms need to specify the target market at a sufficiently granular level, and they're not doing that, or they're burying it in small print and using dark patterns to get people onto mailing lists. There's definitely an agenda across these firms to capture data.

Pat: There are two sides to this. As soon as you start capturing health data, you move into a higher tier of GDPR with much tighter restrictions, and if you're capturing health data without proper permission, you can be in real trouble — that part is enforced. On the other hand, these firms genuinely need some health data to determine if their product is a good fit for someone. So ideally, up front there'd be a privacy statement: "we're going to collect personal and health data from you in order to recommend a suitable product, and for no other purpose — tick to proceed." That's fine.

But if the actual flow is: "£6 a month, £100k cover, tick here to proceed," with all the marketing and privacy checkboxes pre-ticked, and the first question is "have you had cancer, what's your age, any pre-existing conditions" — and then they tell you they can't help you, and a week later you get six follow-up emails pushing other products — that's completely illegal.

Russ: It feels like that. I had a phone call yesterday too — speaking of unwanted contact, I seem to be getting more cold calls than ever. Almost every week, to the point that if I don't recognise the number, I just don't answer. People don't really cold-call like that much anymore — everyone's on WhatsApp or social platforms. So unless I'm expecting a call, if it's out of the blue, I ignore it, because they'll find another way to reach me. It feels like one of these sneaky routes where they've captured my details.

Pat: I use the iPhone's call-screening feature — whenever someone not in my contacts calls, I get a voicemail saying "Hi, it's John from XYZ Limited calling about this," and it shows on screen who's calling before I decide whether to answer. If not, they just leave a voicemail.

Amelia: I didn't even know about that.

Russ: That's cool.

Pat: Android has it too. It's a great feature — it does confuse some less techie people at first, but it's genuinely useful.

Russ: I've had someone calling back repeatedly asking, "what company do you work for?" I kept saying I'm not going to disclose that, and he said he'd just keep calling until I told him.

Amelia: Wow. That's one tactic, I guess.

Russ: Definitely borderline illegal, trying to get me to fill out a profile on myself over the phone. After a couple of calls like that, I've just stopped answering unknown numbers altogether.

Pat: What you should do is say "I'm busy, can I just get your name and email address," then issue a Subject Access Request, which by law they have to respond to — it'll probably cost them two or three days of effort pulling together every piece of information they hold on you.

Russ: This is the other thing — people are so time-poor, they just want to get on with things, and these firms play on that. They know people don't have time to read complicated opt-out messages or go through a Subject Access Request process. People barely have time to fill in the form or take the call — they're busy with work, with kids.

Pat: Ask Claude to do a Subject Access Request for you — two minutes.

Russ: Yeah, I think people just want to get on with their lives and find the path of least resistance. These firms play on that — you go through the process, your details get captured, and you just want to tick the box that says you've got cover in place. If you asked most people what's actually in their life insurance policy, after a few years they'd probably struggle to tell you without digging out the handbook — even though that's exactly when they'd need to know.

Amelia: Shall we move on to our next point, which we touched on earlier — complaints procedures?

Russ: Yes — this matters a lot under Consumer Duty, particularly for vulnerable customers. Some people may lack confidence managing their finances, or have low literacy or low digital skills, so if they're having a problem with the process or communications and need to complain, the complaints process needs to be genuinely accessible. As we've discussed before, some people feel more comfortable speaking to a human. One of these websites made it extremely hard to do that — clear dark patterns in the contact journey, the kind of loop you've probably experienced in retail, now showing up in life insurance.

On the comparison site, Firm B, you eventually land on a page that says: "Should you wish to use an alternative means of communication, on request we are happy to correspond with you by telephone." That's the dead end if you want to speak to a human — genuinely surprising for a well-known site. If someone has access needs, low confidence, or low digital literacy, this assumes everyone can complete online forms and emails, which isn't always true — especially not for vulnerable customers. At minimum, there should be a phone number or a callback form; instead it's vague, and doesn't even commit to giving you a number or calling you back.

Screen reader issues also came up on the complaints form — no easy way to jump between sections, missing alt text on images, so screen reader users won't know what an image is. Real shortcomings on such an important form.

Amelia: Do we think firms are making it this hard to reach a human on purpose, or is it just poor design?

Russ: In the case we looked at, I think it's deliberate — maybe to reduce admin costs, or because they don't want the responsibility of handling the call and whatever follow-up comes with it. But for a vulnerable customer, or anyone wanting to discuss a product before purchasing, speaking to a human should just be a core part of the journey.

I can move straight into the fifth point too, since we're already on screen readers.

Amelia: Our favourite topic.

Russ: This is a recurring issue: the complaints page on one firm's site has no H1. If you know anything about building a website, that's one of the most important semantic elements on a page. Imagine someone using a screen reader who's just clicked a link — the H1 is what confirms they're on the right page. Without one, wrapped properly in an H1 tag, the screen reader can't tell them where they are, unless they work it out by going through the form fields. I tested this myself with Mac VoiceOver. I genuinely can't understand how something like this gets through QA on such a key page. Hopefully Consumer Duty as a framework pushes firms to get this right.

Amelia: These accessibility issues keep coming up across all four teardowns — it's been the biggest recurring issue. Why do we think that is?

Pat: It largely comes down to not taking it seriously enough, or not understanding it well enough. To get accessibility right you need a team that prioritises it from the start, across both design and development. It's a fairly technical, and sadly still niche, part of the digital world — it shouldn't be niche, given how important it is, but it is. Plenty of digital teams design and build a site that looks visually impressive and ticks the client's boxes, maybe run a quick accessibility check at the end, and still miss a lot.

Then, as the site evolves and new pages get added, you need to maintain accessibility over time too. That missing H1, for instance, could easily have been introduced later in the CMS by a junior who didn't realise it mattered, or chose the wrong page template. So there's a requirement to get it right during design and build — because retrofitting accessibility to double-A standard is difficult — and then to maintain it for everything built afterwards: alt text on every image, an accessible tone of voice, proper page structure. It's genuinely difficult to sustain, which is exactly why it's not taken seriously enough, and why it fails more often than not.

Amelia: Do you think AI will improve this?

Pat: Yes and no. If you point an agent at any website, even an inaccessible one, it'll summarise the page and help you navigate it — so AI does mitigate the impact somewhat. But a site with genuinely good accessibility will be easier for AI to parse too, so it'll naturally surface more in AI answers, and an agent pointed at it will read and use it far more accurately. A fully accessible form is easy for an agent to complete; a form that looks fine visually but is a mess underneath — broken labels, missing structure — will take an agent twice as long, and that frustrates everyone, not just people who need accessibility features but also users who just can't be bothered to fill the form out themselves. So accessibility is still going to matter for the wider user base, not only people with permanent or temporary access needs.

Russ: I think AI is genuinely part of the solution here — either through people with access needs using AI agents to interact with these sites and having the agent fix the issues and relay content back in the way that suits them, or through an "augmented" AI layer that fixes issues for the user directly as they browse. That would be exciting — the AI reads the page and returns a properly accessible version tailored to that person's needs. But really, it's still the firm's responsibility to make sure AI isn't quietly papering over gaps they should have fixed. I don't think that responsibility for having the core, accessible version of the site goes away just because AI is helping. It'll be interesting to rerun these tests with UX Duty in a few years and see how many of these findings still show up.

Amelia: Sounds like our podcast episode for three or four years' time is already sorted. So what's the overall summary, Russ?

Russ: Across all four episodes — twelve firms in total — we've found nearly 600 findings. The common theme across bank switching, robo-advisors, mortgage brokers, and now life insurance is loud headline offers that are often close to meaningless, because they apply to a narrow, specific situation that's hard to identify — so it's unclear whether you're even eligible before you start the process. There's a lot of hidden fees too — one mortgage broker required a no-obligation meeting just to find out the fees, which doesn't feel in the spirit of Consumer Duty; it's not transparent what you'll actually pay. Dark patterns around data consent were a consistent theme too — not a surprise going in, but surprising just how common it is across the teardowns, and it's an area that clearly needs improvement.

Complaints processes were missing or buried in the terms and conditions in 10 of the 12 firms, with no clear, or no, route to speak to a human. And the biggest finding overall has been accessibility — of those roughly 600 findings, nearly half were accessibility-related. Breaking that down further: 82% of those were related to screen readers, so firms focusing there would make a huge difference for those users — missing alt text, missing landmarks and headings, missing H1s, all of which scrambles the page for screen reader users. 10% were keyboard-only navigation issues — content hidden from assistive technology, for users who don't navigate with a mouse at all, which we've also seen in usability testing for our own clients. And 5% — lower than I expected — were firms failing minimum colour contrast requirements, which matters not just for people with visual impairments but situationally too, like using a site in bright sunlight or on a poor display.

That's it in a nutshell. I'll turn this into a couple of articles with specific examples, which you'll be able to read on the website. We went into this without knowing what the common themes would be, and looking across all twelve firms, some really stood out. Hopefully this is useful, and if you're a regulated firm thinking about Consumer Duty, you can use it to check your own website for improvements.

Amelia: And of course, there are three other episodes if you haven't caught them — go back and listen. Pat, any surprises for you over the past four weeks?

Pat: Sadly, not really. I think the real surprise was how prevalent this rule-bending for lead generation is within the regulated space specifically. Dealing with a lot of clients here in Guernsey regulated by the GFSC, regulation is a fundamental part of day-to-day life for them, and they take it very seriously — GDPR included. So I was surprised to see so many, mostly UK-based, regulated firms flouting Consumer Duty and GDPR so openly. In my experience, regulated finance firms take regulation seriously — but from a marketing perspective, some of these firms are behaving like an unregulated e-commerce or travel company that has much less risk of being caught. So yes — interesting, and definitely eye-opening.

Amelia: It has been. Next week we move on from Consumer Duty anniversary month to the next big regulatory moment of the quarter — the EU AI Act. What does that involve?

Pat: The EU AI Act was proposed back in 2023, in response to concerns about firms using AI in ways that could harm ordinary people's rights. It categorises AI uses into risk tiers — high, medium, and low. High-risk uses include things like credit scoring, insurance pricing, and CV sifting — all things that are used extensively today. I'll have a good rant about AI in job interviews when we get to that next week.

Some uses are banned outright — for example, using AI to analyse someone's face on a video call to infer their emotions, or for social scoring. High-risk categories include credit scoring, insurance pricing, and CV sifting; low-risk includes things like using AI to help write emails. Penalties can reach €35 million or 7% of global turnover, so it's significant.

It was originally due to come into force this year, but enforcement standards weren't ready, so some of the high-risk provisions have been pushed back to 2027, to give businesses time to adapt and give regulators time to build a proper enforcement framework. It should be an interesting topic — and worth noting the UK isn't in the EU, so the Act doesn't apply here directly. That raises an interesting question: does the UK become a kind of AI Wild West with fewer protections, or does lighter-touch regulation let the UK move toward the benefits of AI faster than the EU? That'll be worth debating.

Amelia: Plenty to get into — that's next Friday. Thanks for joining us, and we'll see you next week.

Russ: Thanks all.

Jargon Busters: Short Selling

Amelia: This week's term is short selling.

Pat: I know what that is — go first, Russ?

Russ: You're putting me on the spot, but I believe it's betting on a stock price going down. Thinking about bull and bear markets — in a bear market you can actually make a lot of money short selling, because you're predicting the price will move in the opposite direction and positioning for that.

Pat: The way it works: you agree to buy shares in a company on a future date at an agreed price. The risk is that the price ends up higher or lower than what you agreed. Short sellers spend days, weeks, or months researching companies they believe are performing worse than they're presenting to the world — like investigative journalists, digging through public accounts, history, and leadership decisions to build a picture of a company that looks healthy on the surface but is a mess underneath. Then they bet the share price will fall, sometimes to the tune of billions. There's a famous short seller who predicted the 2008 mortgage crisis and made huge returns almost overnight when it happened.

Russ: That's the story in The Big Short, isn't it?

Pat: That's right, yes.

Russ: Definitely worth (re)watching — this is reminding me I need to.

Amelia: The official definition: betting that a stock price will fall by selling borrowed shares.

Russ: Borrowed shares — that makes sense.

Pat: I didn't quite get the mechanism exactly right, but I was on the right track.

Amelia: You had the essence of it. More Jargon Busters next week — and next episode we move from Consumer Duty anniversary month to the EU AI Act. Thanks for listening, and we'll see you next week.