Financial Times' Jon Slade on using AI as a catalyst to evolve, with quality journalism as the foundation

Hello and welcome back to the publisher podcast from MediaVoices. I'm Peter Rusted and this season we are bringing you the best sessions from the definitive AI forum for media, information and events which we held with flashes and flames in London at the end of last year. The episode is accompanied by a write-up of the key takeaways which you can find on our site vioices.media or get delivered straight to your inbox every Tuesday by subscribing to the publisher newsletter which is also linked from vioices.media. Before we get to this week's interview I just wanted to let you know that from the end of February through March we'll be running a series of AI-focused masterclasses online. These are the masterclasses we first hosted in person at the AI event in November. They're all focused on how to get the best from AI and your publishing business from Messing ROI to optimising engagement and content discovery. They'll run every Wednesday from the end of February and you can book now at vioices.media/masterclasses. Using a discount code mv25 to save 25% on a 49/99 ticket price. Now back to this week's episode which features John Slade, CEO at the Financial Times, speaking with CDelta's Chris Stunkin about how the business information publisher is taking on the AI challenge. John talked about how he sees AI as part of a larger disruption. The forced publishers back to the fundamentals of strong journalism and strong brands. Along with the development of new product he spoke about the FTs approach to licensing and the brand's long-term focus on direct corporate deals rather than licensing to LLMs. He'll spoke about the role of the FTs archive in creating predictive intelligence products. Chris began by asking John about the challenges of the AI era. You're recently installed, congratulations as a CEO of the FT. That still sounds weird by the way. I think John Riddings' kind of reign, John Riddings' 20 years was probably during the era of paid content for the FT. I think I said to you before, you're right here starting as the CEO who's going to take over just as the AI generative era begins. Is that how you think about your challenge at the FT? I think that's probably not too far off. I think we have a fantastic business that we built over the last 20 years. The paid content side of that is the core. It's what drives 55% of our revenue now and growing. The other side of our business are strong as a consequence of that. Our ad business is good because we've got a good subscription business. We're in a good place and I feel very fortunate to take over from John with that strong base and foundation to build from. I think AI is one of the challenges that we have to get into in the next era. I sang to Colin actually earlier when I was applying for the role. I felt like I needed to have a sort of a five point plan and you can't be a CEO without saying AI is one of my four point plans. I've actually revised that a little bit as we're going into next year and I think disruption in general is a far wider issue that we have to deal with as an industry and at the FT specifically. AI is obviously a very significant component of that but we're also dealing with the changing reading habits of younger readers, for example, which I think is substantial. We're dealing with an assault on truth and trust and facts which come from the centres of politics in the world and that's a massive disruption too. And of course all of the other global turmoil that we're dealing with. However, within that AI is a very big chunk. To some extent I think we go back to first principles. It means investing more in our journalism and in original journalism. It means investing more in telling the story of the brand and neither of those are technology solutions but they are a response to a technology issue. But it also means leaning into technology whether it's experimenting with it for our own internal use or developing new lines of business or developing new customer products. And I think the licensing conversation that I was just listening to there is a really, for us actually a quite exciting opportunity for developing a new line of business. Just on that I hear a lot of talk about the pitfalls or the upsides of licensing to large language models or to big tech companies. That's, we do some of that, we've done some of that, it has some upside to it. It has some risk to it, certainly not a perfect model. But we're much more interested in the direct licensing we can do to corporates. We have an excellent licensing business which is built around human licensing. We also call that subscriptions. But we license access to the FT through our corporate business. And we're now building into that machine reading licenses as well. So you can take an access license to the FT where all of your people can read our journalism. But you can also take a machine reading license where you're not just getting the 2D artifacts of the article, but you're getting all of the data behind that as well. And you're getting access to our archive. And there's much more control for us in that business which allows us to be much more confident about how it's being used, how it's being surfaced, and to set controls around usage. And that's a close to, as I'm a bank, let's say. I'm a legal lawyer. You're a legal company, you're a communications business, you can have a human reading license. And as an adjacency to that we can build a machine reading license as well. I think that's a very exciting area for us to develop. I'm much more sustainable business around. I think large language model big tech licensing, I want the same thing for Christmas as Tammy does. I'm not sure which Christmas that's going to arrive. No. I don't think it'll be this Christmas. But if we can build some kind of controls into that great, I am not hugely confident that there is a sustainable business around that type of licensing. I'm much more confident there's a sustainable business indirect to business. So you've got some interesting positions, right? You're a long-term Google partner. You are an open AI deal partner. You're a perplexity litigant. We are not. No, NKR parent company are injured. I know that. You are somehow remotely involved in litigating perplexity. So you're kind of a mix. You're doing deals, but you're also working with partners. You're in and you're out. Can you just explain how you come about that? I think people will be really interested in when you think about literally signing those deals. What's the kind of, what's the decision-making process within the FT about thinking about that? Well, we've learnt a lot as we've gone through it. And we are probably two and a half years since we signed our first deal. And we have a three for reasonable-ish deals that we've signed. And I think each time we've signed one, we've learnt more as we've gone. They are of different scale. They include different parts of our archive from lots of it to not much of it. They include some of them include live feed, some do not, some have a delay on them, some don't, some allow for surfacing, some don't. There's quite a spectrum of respect to it. Predominantly we're looking for the maximum amount of control that we can get over the use of our content. We are always looking at a very common thing in all of those deals is how we got what we call the big red button. And the big red button is the thing that allows us to go, we don't like what we're seeing. We've got a hope that we have the transparency that we've agreed in the contract but assuming that we do. If we don't like what we're seeing, we press the big red button. Now that doesn't take the salt out of the soup, so to speak, but it does stop access to our archive. It does stop any further surfacing of content. And that allows us to feel relatively relaxed that if we don't like it, we can cut out. So not everybody that we've spoken to has agreed to that. And so we haven't signed those deals. So as a first order, if we can't get out, we don't like it, then we're not going to do the deal. Then we're going to look at the money that's attached. And that's complicated and subjective. And funnily enough, we think what we've got is worth more than what our buyers think. And our job is to kind of talk that out. So we're in a privileged position at the FT because I think there is an implicit sense that FT journalism is of high value and of high worth. But putting a number to it has proven difficult. I would say I'm broadly happy with the kind of numbers that we've achieved. They've certainly been useful. But back to my earlier point, I don't think they're sustainable. It's not a business. I promised in the spirit of transparency. I said to John, I wouldn't ask him the exact commercial nature of the deals. So how much were they worth? Well, the first one I'm going to live. So we're just interested in that concept of three or four deals and each one slightly different. Is that sort of an optimisation? Is it the last one you think is much, much better than the first one? Or has that been sort of horses for courses? Good question. It's probably a bit of a mix of that. I think we've learnt as we've gone through the deals, which aspects we need to push harder on. So liability, for example, is a kind of non-negotiable. We have precedent from previous deals that we can feel more confident that actually the partner we're dealing with is not too dissimilar to the previous deal. If it was OK with them, obviously, why shouldn't it be OK with you? So there's precedent builds into that as well. But yeah, to an extent, it's been opportunistic. And let's just divide the licensing deal with Anne, it's a straight deal. I've sold a license. Quite a lot of the deals that I've seen have been, I've done a deal for a licensing of the content. And I'll take on a technology partnership, tools, newsroom training. It's a package deal. You've done those kind of things. Wow, ond wow Aun yn marnol a'r ivoli o sembliol o goneyه â cynsu wedi'n ran eradwydd… …a'r трá cosplay wedi hon f receiving i castar Pos Roeddwn lwybod sy'n felcharac yr ordiniodwch mental McLeodd Pom oedd warg y gyfer fy itindante fe. Tre. Om bell angen arall agdwyr eraill ddude diwed yn arwedig neithu yna yn felwyr. Mae'n d調 Yn sraentaol génym gorillau reconstructatolol neu galli ei ac teasynio testediaeth. Chineynydd y Eye capigol o'n reichen haw yn meddu nad hwndru beth unMoment yn beth ei chi wedi ei gallu ddalwbod. Mae'r through 온 na mewn meddwl a enfauddwch, yna g adaptive r食d beредlwst er no whewar Essen Zingushi ddum capable o'r Brynorg bloegol un nad un w Weithynfiol ddan Agutianosovar Wh societ yn y char73 o sgbyddiad. Thym dy nyd o'n teth ar y قymnaeth boomh 95 i ein cymraeth Cynyn ei rel Gabriel a'r ysgwch chi'n gydyn yn y cymraeth Cyny ei ac ysgwch chi'n gydyn yn y cymraeth ac mae'r ysgwch chi'n gydyn yn y cymraeth Cyny ei ac mae'r ysgwch chi'n gydyn yn y cymraeth after depending onces with the fresh Kentucky Werdzuli un�c i tekau agau phenal efo rhyngio ac ourin? Mae'n panol cael dyna agau dystaleos am amg dag argyn nhw, OK,ämän ran dych yn sy'n siop. Dbell i yw ag fo sy'n gweithiau iawn gwerth o reduce, neu gwen Dwirlo 과'r saffreitir mewn, ac mae iawn gyfer cystag siwryw ydo ei yw yn blafysio i dder ekwbethau gyfer yn dystlawn Visa�ol, gennym wedi wneud eu eu gwerthau gwntau, period pawf arall maybwydd ei os nhw мет Guardians a myfem yn g crafts i fl gymidd. Ond yna ond ddef approach dar fŵr y fıyor a f זodaeth rydforde crad, y gymidd y Tikbrid? Dkinserort cryro hyd meff sonygygu grywch at g倒 euu rynu diad i makesgef i y sefymd fiberwyrredd anedd toud ac mae felly beên fo ddu genni wh erm oes. Lohon i dda ddarwol i wrth gael fel arall roedd mweileig o'n ffe��raged i d�id i yn g locking staff am symrankaullen a tyddio o'r efo kotwn sydd cywllag optimum. Voin yn es gísm yn cyfan. To'r garr freshwater o 'ag nylon Pig. I'r gydyn yn eroeddumaith apthдиadau' o fei'r mapaóg gen sydd o eto dod i'r gym就kan, hwyr sięg bus Sidforddd yn rais yn hi. D superhero o grfo hiculidwch. Chydaw y pudna sglorodesaedd rotate. Unos ni'n rhwng i gwasil ceir uz ong. Yng Nghymru, unor fwy barfo reiillagin hype. fått yny gwir unaен d принципе, yma fwngdd bumrydd syddarnol gan nhwyn nhwh bound eichraeth. Gwyr i'r gwaith y gwaith. Gwyr i'r gwaith y gwaith y gwaith. Gwyr i'r gwaith. Gwyr i'r gwaith. Gwyr i'r gwaith. Gwyr i'r gwaith. mark, Markerd Gly Oxfordd y Cre Chinese Wr Alexa, coer 위d yn rydych chi 'dyrch yn RoedrayngByny' fel Whomayryd gweld seig gan am yn wirwn ddwyntarebyn mae'n i'r sierach heddiw drwynaeth gan somt Catcha ymaau mewn lied אוizhygn ein hyngin i ddefnyddeg gドンstel Deputy Godd Alo'r maegolio總as wediということan pobl triggering Natu ddol Dwi-breda gweld i gwertyg eitol unusual Mae этом �raisывайтесь etyn glor, tani rydyd yn tro yn siarau pewn lot ar ymfestryeth agoodd I hearthiriau ednaை V mynd ac ben Yffield dyna sy'n rgliad de亜 yn openillatfo dr Arm biraz yma meddinea zas anion a'r l appetite i callar Lyll挸� neu b ihnen lehidiaeth yn ni sydd pythywidd m addor a'r maldind prostu deill hon honfedna, prosi ddod yn прирfwyrfa a garaffaethu unes gynharf erwt夷 c Des Nemwysiaeth i'r ll調 yazd, yn prasod espisaeth eu ceffaethu go starol- roedd o-Danarioll, rydol dderwyb y bithераeth ar hy- ni cw точно sydd Thoughted OA lawyers hymid o drafllr yma llawer o gwelir jar yn gallos i gael fel cofwyllol reigauau gyfer awer yn mladed ewn sraeg yn y guffppir i'r cyffiran i gwasín Mae'n dod gygindechrau wrth gynhylopersol childcare gwneud podelost丸 gèindu, gydwysig i gofir gyfer y pleim還odraho y ffigio fy語wyr wrth gydwyd beth i j Gunnyddwr, gydwysig i gofer, rydwysig sydd anag ymwyr s daughters' from the roGeEnd down byg chi ywi wrth g이� deboli unig at yr nearly unag yad wedi larodraedd ymgylch nehosrud гethili, it'r thrES ar fy hwns cael pom y fyw fi gyleidodd ag erffオ gyfer cyfiosu'r yr wyth scaho Stargwerdych fuswn o ffordd o modedorgMom. bun yn rodan dddull, i Risew Cyfrutar nhw'n werid gw yearn gweith像 a therapyfa y нем Nghyw Fawrddol pan dataset orastar gwaith scalebw'n dod wo ym sûrach ddowswch i'r Fiwrddol mewn hyd yn bydd yng Nghyw ongyny, kaethy'n bywyr by'r synth. Fawrddol yna Rhomo nesrau'r rhomo. Fawrddol yn cael bedwybethedd ac了. Mae'n dod oesrau'r rhomo, ac mae'n dod oesrau'r rhomo, ac mae'n dod oesrau'r rhomo. Mae'n dod oesrau'r rhomo. Mae'n dod oesrau. Mae'n dod oesrau. Mae'n dod oesrau. Mae'n dod oesrau. Mae'n dod oesrau. a'n eu gweithio. Cael ei chwyleipaeth o cho wedi hebun yw ar un riffadu bon sharperaf hwysabsifad dedication t pollingau o'f exaggeranteし i edry ar chi dobl meinat. U wedi'i wneud ddi centuryd an islandai bod faci, ��fysffen wedi einไปio'tio at bawbraf��, sy'n ond dim fyralds y caeldeaf losimo a hunar of fod, unir enghafal yn akhir isfewci; elu wedi yn pedig rydwys gweld EspeciallyonAndai'r Wrang hyfr�au'n hyfr comeeb officiしad wedi mudd gor Fehlerrys yn manifestiad am wayna fue gynhini Tipenderos. Cael gratulationsren wet n Eden ar ddwyn e Sinnem ond a fellymlen nawr gapaf ond eich geil i e cael dorol fel cyprofall liflu ei unnecessell tua nest yn a faltos modeol fel Herópooryn A mae間iad erectio ble oedioel yn dweud Edward Neetova freeze a feddwl ond a dde syloer wedi pethrau gallu gyda'r i'n pethrau stabil�� da mae fr accents yma cynnyddians remarg絵 o'n pethrau? Mae hyn eu yn y cwedd Yn wneud acles gwneud gorillan actionsaeld i ni ein niegu o bob el wneud offline naill am ypwch molecule lefer, un bod mewn law—mwyr囃r adryr yn flau. Naww'r maenig hon pwkerpryddeid yn cefdwcheg fwybryd Anglwyr cyf壓 yn gallóng yn hislene b locations y postul waywn organisations cór ter 떠� uwag. Hwybf mewn feddwl y pan didaethoc mewn lebypoeluch wedi bennydd odny felt gyms Ewiraeth YsMary, hon gw wisdomd a a wedi Line'n hon qu Watches system a wedi'r pros iawn callangol yn Semb heatwwrdd 그래도 hblyna golygu? Ys foall ei gyffrwwch pan ar fyddyl y cyfoddiol yn daeth cym Jumpers Rydych chi, am��은 eisiau rel yn'm, eisiau felly flyr yn g clerfryd, ryn yn cynnodil﹐ hyn ein rebytatau, yn doch mae'r cyらい ar yszym awn sylwsad ac rhyION %, mae fy'r sousfnunga, i d� Considering eraill yn hollfa acor, Realm i'n teres ar façon gen y gyfan ei ffightu d am yna newydd yn g EMHT i cynnodatau Dawn yn Gwyli swydd ydna angen yw'r cyfod yn gysylltu, gynny, mae'r cyfod yn gynny, ac mae'r cyfod yn gynny, ac mae'r cyfod yn gynny, ac mae'r cyfod yn gynny, mae'r cyfod yn gynny. Mae'r cyfod yn gynny, mae'r cyfod yn gynny. Mae'r cyfod yn gynny. [Mu'n ddolch]