2025 Recap and 2026 Predictions!

19 Dec 2025 · 31 min · 14 chapters

Ask about this episode

Ask anything about it. ChatGPT or Claude reads this page and answers with the times it was said.

Connect VO and ask about every podcast you hear, including the moments you saved. Add to ChatGPT · Add to Claude

In short

SlatorPod year-end recap of 2025 and 2026 predictions for the multilingual voice AI, audio AI, translation, localization, and language services markets.

Guests

Esther and Florian (SlatorPod hosts). No external guests are discussed in this episode.

Guest/host backgrounds

Esther and Florian are Slator researchers/industry analysts tied to Slator’s market research (e.g., Slator 2025 market recap) and SlatorCon London.

Key claims

Multilingual voice and audio AI stayed a major investment hotspot; 2025 showed rapid expansion from LTPs toward full-service “LSI” offerings; 2026 will bring continued consolidation and more partnerships; voice/audio will keep driving disruption.

Notable examples

Synthesia $180M Series D; ElevenLabs $180M Series C and Meta “language technology partner” program; Anthropic study on translators for AI QA; Microsoft healthcare voice AI in Spanish; Amazon AI dubbing pilot (12 titles); RWS acquisition of Papercup; YouTube AI dubbing rollout backlash; TransPerfect acquisition of Unbabel; 11Labs “Productions” managed service; Middlebury Institute/College translation program closures; Cisco acquisition of Easedubs; 2025 public-company financial updates (Straker, Appen, RWS, AMN, Teleperformance).

Written by AI. May contain mistakes. Listen to the episode to check what was said.

Chapters

Tap a time to open that second in VO

2025 Market Recap Overview

0:45 to 2:28

Discussion about the Slator 2025 market recap report and its structure.

“So we're going to go through this and we're also going to do a quick predictions and outlook segment at the end.”

January Highlights

2:28 to 3:05

Key funding stories in January, featuring Synthesia and Eleven Labs.

“So I'm actually curious what you're thinking about what's important for January, Esther, January 2025.”

February Developments

3:05 to 4:38

February highlights including Meta's language tech program and an Anthropic study.

“And secondly, that Eleven Labs had also raised$180 million, but this time in a Series C round.”

March Insights

4:38 to 6:04

Discussion on McKinsey's AI report and Trump's language designation in March.

“like a runner up or something like that, because I'm also going to mention too.”

April's AI Innovations

6:04 to 8:06

Highlighting Microsoft's healthcare AI and the Legal Services report in April.

“And that kind of caused some ripples, I think, further on in the year as well.”

May's AI Dubbing Pilot

8:06 to 9:07

Exploration of Amazon's AI dubbing pilot for Prime Video in May.

“So one of the top stories and my kind of pick also from the recap is Amazon's AI dubbing pilot for Prime Video.”

June's Language Tech Developments

9:07 to 10:19

Discussion on RWS's acquisition and tech updates in June.

“I have yet to watch an AI-dubbed Amazon Prime video movie.”

July's YouTube Dubbing Rollout

10:19 to 11:58

Exploring YouTube's rollout of AI dubbing and its reception in July.

“I was going to also mention the AI dubbing rollout by YouTube.”

August Acquisitions and Innovations

11:58 to 14:00

TransPerfect's acquisition and OpenAI's hiring trends in August.

“And I'm going to do something that I don't think we talked about but I have to go back to May because we forgot something We left something out, some big piece, which was our new market framework that was launched.”

Localization Manager Trends and Updates

14:00 to 17:00

Discussing the evolving role of localization managers and industry developments.

“And the next thing they do, they hire a localization manager.”
Show all 14 chapters

Financial Performance of Language Companies

17:00 to 24:40

Reviewing financial results and trends from various publicly traded language companies.

“Complexity also started looking for a localization manager and remember that they basically just copy-pasted OpenAI's job ad.”

Predictions for 2026 in Language Services

24:40 to 28:00

Making predictions about the future of the language industry and emerging trends for 2026.

“and their businesses are going well and growing very quickly.”

Market Consolidation Predictions

28:00 to 29:13

Exploring trends in market consolidation and acquisition strategies.

“I think we're going to see consolidation happening at those levels.”

The Rise of XLIF2 in Localization

29:13 to 30:35

Discussing the potential of XLIF2 to replace TMX in translation management.

“thought I threw it in there given all the macro stuff we've talked about that XLift2 may replace or at least strongly compete with TMX in just general localization management.”
Hear the part that matters, and keep it.Open this episode in VO. Double tap your headphones to save a moment as you listen.
Get VO free

Transcript

Automatic transcript. May contain errors.

0:00Florian Faes:Multilingual voice AI and audio AI continue to be a hotspot for startups and investors.

0:13Esther Bond:Hey everyone, and welcome to a very special year-end episode of SlatorPod. Hi there, Esther.

0:18Florian Faes:Hi, Florian.

0:20Esther Bond:37th podcast this year. Shelby, the last one. How are you?

0:24Florian Faes:Pretty good. Looking forward to finding out all about what's happened in 2025 and unpacking that with you.

0:30Esther Bond:100%, and we're relying very heavily on our very most recent research report, which is the Slater 2025 market recap, phenomenal report out this week. 120 pages, great recap of the year and available now for subscribers. And you can also get it as a standalone report. So we're going to go through this and we're also going to do a quick predictions and outlook segment at the end. And before that, quick reminder, SlaterCon London coming up on May 22nd. Tickets available now, so go and get it. So what we want to try to do today is to go through 2025 month by month, and we're going to pick one random piece of news, one random development, one something that we consider to be important in that particular month.

1:18Esther Bond:This is the first time we're doing this, so cut us some slack. We'll give our best, all right? So I guess we said that you were going to start with January. I'll take February and then you're going to do March and so on and so forth. So, it's going to be a very tough one to summarize for any AI you guys are using because it's going to be a very eclectic choice of individual pieces of news or developments that we're going to go through. So, just very briefly before we get started, I mean, the report itself is structured as basically news and insights by month. And then we have another 40 pages of research highlights from our very own research.

2:03Esther Bond:Very, very great. So yeah, it's about 40 pages. So you get a lot of intel that is otherwise in our standalone reports. And then we end on 30 pages of thought leadership, bunch of interviews and things that we've recorded, podcast insights, et cetera. So tons of tons of value in this report first time we publish it. All right, without further ado, let's get started. And we haven't even rehearsed this at all. So I'm actually curious what you're thinking about what's important for January, Esther, January 2025.

2:37Florian Faes:It could be interesting. But yes, this is our, I suppose, super condensed analysis or pick. And I'm already going to break the rules because I have two stories for January. But they do have a commonality, which is the number 180. So yeah, my top picks for January 2025 was firstly the news that Synthesia had closed a$180 million round, a Series D. And secondly, that Eleven Labs had also raised$180 million, but this time in a Series C round. So my picks, if you like, are two sort of fairly large funding stories.

3:18Esther Bond:The question is, have they spent those$180 million each wisely in the past 11 months? We shall see. We shall see. 11 Labs seems like it because they just announced a big deal with Meta, which is a beautiful segue to one of the stories that I have for February, Esther, which is, if I pick February there, Meta launches language technology partner program to source AI translation data. so this built on the earlier efforts by Meta including the no language left behind massively multilingual speech projects that they have and this was basically a call to action for language technology partners to contribute speech recordings, transcriptions, written text, translated sentences especially for the underserved and low resource languages contributors were required to provide at least 10 hours of speech data with transcriptions plus 200 sentences So interesting partner program by Meta there.

4:18Esther Bond:And then, yeah, I'm also adding a second story, sorry. It was that study that Anthropic ran where they looked at 4 million cloud conversations and suggested that translators lead kind of the general adoption of AI as a quality assurance tool. So that was Jan and Feb.

4:38Florian Faes:Marge, well, maybe we could frame this more like a sort of top pick and then like a runner up or something like that, because I'm also going to mention too. McKinsey published its State of AI report in March. We liked this because it gave us a bit of an insight into their view on the role of a translator. And they said that they classified translator as an AI related role, for one thing, which is obviously relevant. And it also said that in the report that the adoption of AI translation technologies had become universal. You know, something that we know within our industry already, obviously validation from sort of this external expert also.

5:26Florian Faes:They also said that the hiring difficulty for translators had dropped. And that went from over 70 % difficulty in 2022, down to 55 % difficulty in 2024. So that was a good report and something to, I think, go click back on if you didn't read it at the time. My second top story that I think you have to mention, because it had a bit of a longer term and still impacting on various areas of the industry, was President Trump's designation of English as the sole official language of the US via executive order. And that kind of caused some ripples, I think, further on in the year as well.

6:08Esther Bond:That was probably one of our most viewed podcasts, I think that year with Deepak Patel from Globa and Peter Argandiza from Argo translation. Yeah, that was a very good one. English as a national language. Does the UK have a national language? Like officially, do you know that? Like codified somewhere?

6:26Florian Faes:My assumption is yes, but I wouldn't know. I don't know the sort of official channels for doing that because I mean, certainly there are other domestic languages like Welsh and Irish and Scotch.

6:41Esther Bond:whatever it is uh island gaelic but then obviously we have i mean our multilingual provision especially

6:48Florian Faes:within the public sector it is pretty extensive um like from in emails to just everywhere all sorts of communication at government level or um sort of council level either like a local level all of that is is pretty much um rolled out in seven or ten different languages that they would deemed to be most appropriate for the local community.

7:10Esther Bond:Also happened in March, Slater 2025 localization buyer survey. Just want to highlight the research. Actually, we skipped that for January where we had the AI interpreting. Speaking of interpreting, speaking of Trump and speaking of the United States and the national language, my story for April is that Microsoft, they had this piece of, I don't know, kind of semi-viral content where they said that new healthcare voice AI enables encounters in Spanish without a translator. And that's after they launched Microsoft Dragon Co-pilot, which was kind of a unified healthcare voice AI assistant combining Dragon Medical One with something called DAX to support clinical interactions.

7:53Esther Bond:Why relevant? US medical interpreting, huge market, right? And Microsoft going after it, if even just a little bit, seems to be a pretty big deal for the language industry. And then in April also, we published the 2025 Legal Services and Language AI Report. There's no month without AI. I can predict that already. All right. We'll jump to May.

8:19Florian Faes:So one of the top stories and my kind of pick also from the recap is Amazon's AI dubbing pilot for Prime Video. So this was a pilot, a dubbing pilot, as it suggests, which used, surprise, surprise, AI and human post editors to localize 12 titles into English and Spanish. So there was a bit of some quotes here from Amazon saying, AI-aided dubbing is only available on titles that do not have dubbing support. And we're eager to explore a new way to make series and movies more accessible and enjoyable. So the pilot itself covered 12 movies and series in English and Latin American and Spanish that they said would not have otherwise been dubbed.

9:07Esther Bond:I have yet to watch an AI-dubbed Amazon Prime video movie. It's also not huge. Amazon Prime is not big in Switzerland. I think they somehow, I mean, Amazon generally is just skipping Switzerland. I don't know why. Maybe because of the languages. All right, we shall move on to June and also AI dubbing. A big story there that RWS acquired the IP of LTP Papercup, language technology platform Papercup, which we had on the podcast way back. And then let me quote from RWS's half-year report that they say, we propose to integrate dubbing and sub-language capabilities into Trados to offer high-quality human-in-the-loop dubbing to media and enterprise clients who expect demand for both scale and quality in the fast-growing dubbing market.

9:57Esther Bond:I think that, yeah, that was the IP because remember the team, the actual team, the Papercup team, they went to Scale AI, the data annotator that was kind of now semi-owned by Meta, which kind of feels like an ancient story already that Meta bought or invested in Scale AI or bought Scale AI. So that was June and that was a, was that a slow month? Maybe a little bit slow.

10:25Florian Faes:Could have been.

10:27Esther Bond:So let's go to July.

10:29Florian Faes:Yeah, July was pretty interesting. I was going to also mention the AI dubbing rollout by YouTube. The trouble is that there's a bit of a spoiler because, you know, obviously fast forward sort of six months and we know how some of that unfolded. But the initial story back in July, I think, was that YouTube was doing this large scale rollout of AI dubbing. Obviously, there was lots of kind of testing and lots of people commenting and reacting to the rollout of AI dubbing on YouTube, on the platform. We also noted that our podcast had been automatically dubbed into more than 10 languages, right?

11:09Esther Bond:Behind the scenes, though, at least they didn't publish it. So, yeah, it was just sitting there for us to click publish.

11:14Florian Faes:Yeah, so you have some kind of control. At least it's not pushing this live without your sort of express permission. But yeah, I think the greatest story here is what unfolded a few months later, which is that there was a lot of pushback, a lot of backlash, essentially, because people were slamming the AI dubbing for being too robotic or even cringey. and I think putting a lot of pressure on YouTube for all creators to have access to this multi-language audio tracks so the ability to upload their own language tracks which had been available to some YouTube creators but not everybody so that was sort of the, I suppose, the latest outcome in that saga which happened in September

11:59Esther Bond:And I'm going to do something that I don't think we talked about but I have to go back to May because we forgot something We left something out, some big piece, which was our new market framework that was launched. The Language Industry Market Report, which included the framework back in 2020, like on the 22nd of May, we published it right before SlatorCon and that was basically our new industry view, which included the Language Solutions Integrator and Language Technology Platform framework and a ton of other stuff. Plus, we celebrated our 10 years anniversary in London with the entire company.

12:35Esther Bond:That was great.

12:36Florian Faes:Yeah. That was fun.

12:38Esther Bond:What was the bar called? Something with sheep? No. I'm not going to remember. No, no, no. The grazing goat.

12:45Florian Faes:Grazing goat.

12:47Esther Bond:It was called the grazing goat. They keep sending me emails now trying to get me back into the grazing goat in London. Maybe next year. All right. Okay. Let's go forward again. Jump from July to August. And in August, and this feels kind of like a throwback to earlier years, TransPerfect acquired Unbobble. I remember the 2021 or maybe in 2018 when we like every other week, we would have some type of big name transaction to report on. So yeah, in August 2025, TransPerfect acquired Unbobble to bolt down a bunch of AI capabilities. We don't want to unpack that too much, but Unbobble obviously, big name in the industry, they have a bunch of proprietary tech, the Tower Plus, the Tower L &M, Comet, all of this good stuff that's now getting integrated into the behemoth TransPerfect based in the town of New York.

13:45Esther Bond:And then also that was the August was when OpenAI started hiring for a localization manager, which I think they now filled that role, which obviously we unpacked at the time. The irony of OpenAI basically being one of the key drivers behind the whole disruption, AI replacement narrative. Not yet. Not yet. And the next thing they do, they hire a localization manager. So whoever got the job, congrats. I do want to add a third story, Cohere, LLM builder Cohere, they raised a big round and then they launched Command-A Translate, which is a major translation kind of offering, AI translation offering.

14:27Esther Bond:And then, yeah, we're very pleased to have the lead behind that project at SlaterCon in September. So what happened in September, Esther?

14:35Florian Faes:In September, I will talk about 11 lads again. So remember back in January, they raised 180 million. Well, in September, they launched Productions. So this was a managed service providing human edited multilingual content. So this is complete with permanent pricing, expert in the loop workflows, etc. This is interesting, I think, specifically because it signaled 11-lips expansion from pure tech, i.e. under the bounds of our new framework being an LTP, moving into becoming an LSI, a language solutions integrator with sort of full service offering.

15:21Esther Bond:That was big news for us. I wonder if they realized that they were transitioning from an LTP to an LSI by adding productions, but well, it is what it is.

15:31Florian Faes:It's fantastic.

15:33Esther Bond:And then I think we had one more story. If I'm looking at the notes here that we made, we did put together a few notes. I think the Middlebury College one is something we kind of want to briefly mention as well that is big news in academia yeah i think academia but

15:46Florian Faes:also just sort of in that part of the world by that i mean sort of you know california silicon valley localization globalization world i think you know we've made the comment before that there are many many uh industry professionals um i think you know very influential at this point in their career having gone to the Middlebury College, Middlebury Institute, right? So I think it's kind of sent a bit of, yeah, not shockwaves, but some concern and obviously sadness, I think, among that cohort that the graduate programs are closing.

Read the full transcript

16:26Esther Bond:Yeah, so that was the kicker, right? That the graduate programs were closing, like all graduate programs, and that included translation interpreting localization program. So let's see if that really is the final decision. I think they kind of, I'm not sure, you can't say they put it up for sale, but I think somebody could take this program and run it out of it. But yeah, anyway, sad news. A little premature because localization managers, drumroll, are still indispensable as we found out in October.

17:00Florian Faes:Not yet.

17:00Esther Bond:Because after OpenAI, perplexity. Complexity also started looking for a localization manager and remember that they basically just copy-pasted OpenAI's job ad. We had that discussion. That was probably one of the funnier episodes that you just take OpenAI's job ad for the localization manager, copy-paste, add a little bit of, I don't know, some AI content on top of it and then off you go. But if you're listening and you're the localization manager at Perplexity, give us a shout out. Standing invite to come on the podcast and tell us a bit more about what that means to be a localization manager at Perplexity.

17:43Esther Bond:Also want to highlight one more story. Because I think I got it wrong. My take was not great when I discussed it. Maybe not on a podcast, but on some other forum, maybe at remote. So Cisco bought EaseDubs, like a language technology platform. E-Dubbs, you're right. Yeah. And I didn't realize at the time that, I mean, so Cisco owns WebEx.

18:06Florian Faes:Correct.

18:06Esther Bond:Right? So then it makes all sense. Like Cisco, okay, so they have WebEx, WebEx needs to be multilingual, so you add E-Dubbs and off you go. And it was just that the PR was a little over the top when Cisco said E-Dubbs technology She fills a critical gap in the market, answering the call for truly global accessible AI power communication. Like they're basically acquihiring a small company and then they're saying that's like it's filling this big gap in Cisco stack. It's like, dude, like you could just, I mean, it's Cisco, right? You have a lot of means and resources. Moving on to November.

18:48Florian Faes:Yeah, November, I think maybe I'll go back and pick out a couple of financial results because I think that paints a bit of a picture as well. So obviously we do track the publicly listed companies within the language industry. There is a sort of six-month cycle typically of reporting. So I think in November or there or thereabouts, we had quite a lot of updates coming out of the publicly traded companies. So we had Straker reporting year-on-year revenue declines, but they did secure a renewed IBM contract that relates to translation AI in Slack, IBM's partner network, and the joint development of small language models.

19:36Florian Faes:Then you had Appen out in Australia saying that there was a volatility in USLLM project timing and volumes, and that that was the primary factor behind its weak performance. But they pointed out its China business had remained comparatively strong. So really a mixed bag so far. Then we had RWS here in the UK telling investors that revenue was slightly down and profits were slightly up. And the response from investors was pretty negative with the stock going down more than 15 % on that news.

20:16Esther Bond:Now it's up though. Now it's up, yeah. I'm just looking at the chart because they did another. Yeah. So that was in November-ish, like late October, then it went down and now it's back up, it went down to like 69. Now it's back up to 88. So it's getting kind of whacked about. So it's up like 20 % since then.

20:36Florian Faes:Bit of a yo-yo ping pong situation. Then over at AMN Healthcare in the US, the language services division remained flat. This, I think, was it's Q3, must have been. So it remained flat at$75 million in revenue for the quarter. They said minutes were up, so doing more volume, but pricing was down. So equals flat, I suppose, more volume, but lower prices. They seem to have maintained revenues on that. And the CEO said that they believe language services remains a growth business, obviously, despite some of the rapid advances in automation, as we have been discussing at other points during the year.

21:23And then Teleperformance, which is a France-based company, but the owner of LanguageLine in the US, they highlighted the division's resilience and said there's a stronger pipeline quality despite a volatile

21:39Florian Faes:US environment.

21:42Esther Bond:Teleperformance, huh? Also, shares really under pressure, like multi-year low. Crazy. December, shall we do December? We have 17 days to catch up on, but in our report we had, let me pluck out three things that we had under product development. So basically one is this, we spoke about our last podcast, the Amazon AI Kindle translation thingy, which I think is a bit, this is big news, huge, huge, huge distribution channel for something AI produced and as kind of sensitive as literature translation. So I really, I mean, we say this from time to time, but I really, really do want to check back in like half a year if anybody's now, or not anybody, if people, if massive numbers of people are now actually consuming AI translated literature, I'd be extremely curious if this is working, right?

22:38Esther Bond:Because yeah, and here we have an experiment. I mean, it couldn't be a better experiment. If that works and people make money off of it, like the authors, yeah, there you go.

22:50Florian Faes:Great use case for AI translation.

22:52Esther Bond:And then I'm not even sure where to start, but Google is rolling out all these kind of real-time speech-to-speech models, but also applications. I think I saw something on YouTube recently where they had a demo with even, I think it was EarPods, but it was off of some type of Android phone. So yeah, live translation, live AI translation. All right, so yeah, because we only have three weeks, so that's it for December, which took us half an hour-ish, 25 minutes. Good stuff.

23:27Florian Faes:Yeah, we tried to be as chronological and as logical as possible, but yeah.

23:32Esther Bond:Yeah.

23:33Florian Faes:The jury's out on how successful we have been.

23:36Esther Bond:Hey, it's free. It's free. We put it on YouTube. It's free. Okay, so we also said we're going to do predictions. So basically before predictions, just quick outlook for, I think it's a tale of two worlds. I think for the, I guess, legacy or established or traditional or whatever you want to call them LSIs, even LTPs, super tough year ahead. Again, you have to rebuild parts of your platform. You have to basically say goodbye to clients that you've been building up for one or two decades because they're doing all kinds of other stuff and experimenting with more automated solutions. So you have to adjust, be super nimble, super agile.

24:19Esther Bond:But then there's also an incredible, there's just incredible activity on the startup side, on new companies, new business models that is very exciting, right? To many, including us. I mean, look at the type of people that we have at our conferences. It's great. I mean, so many people that you wouldn't have met five years ago are now coming and their businesses are going well and growing very quickly. So yeah, tale of two worlds, really. And I think both of these cohorts need to risk, need to kind of run some risky bets again in 2026 and that's probably easier for the startups to do because they have relatively little to lose while some of the established players have more to lose.

25:12Esther Bond:But yeah, risk is something that you should run in 2026. And I think the partnership angle is very important. Also think about, we said, we call them LSIs, Language Solutions Integrator. So you need to partner to be able to integrate things. So partner, partner, partner. There's so much good tech out there, so many good tech providers and other providers that can help you build a solution for a specific client use case. So I think partnerships is going to be huge for LSI's and LTP's also in 2026. But predictions, let's do three each. Okay.

25:50Florian Faes:So my first one is that multilingual voice AI and audio AI continue to be a hotspot for startups and investors.

26:02Esther Bond:I have kind of a similar one and again, we didn't align before so it's interesting. Like my first one was like huge realignment on the LTP side as voice blends firmly into the mix. And also, yeah, so basically new competitors emerging in various kind of established areas. Like we just had 3Play on the podcast and they moved into AI dubbing, right? and some AI interpreting companies may enter AI dubbing and dubbing may enter AI interpreting and so on. Basically, this is against the backdrop of just voice and audio demand growing very strongly. So I guess we agree on that, that voice and audio are really going to be a huge thing in 2026.

26:48Esther Bond:All right, your next one.

26:50Florian Faes:I predict there will be an announcement of a super agency sale.

26:56Esther Bond:So somebody is going to buy a super agency.

26:59Florian Faes:We always predict this.

27:01Esther Bond:Well, it's overdue. At some point, you're going to be right. Yeah. So, all right, super agency is going to transact finally in 2026. Similarish from my side, number two is one of the bigger LTPs or one of the biggest LTPs is going to go IPO. Somebody needs to go finally IPO, some big one.

27:22Florian Faes:What about where would you care to? Care to bet on where they would like to.

27:29Esther Bond:I think it's got to be in the US. US? Yeah. Sadly. I think, who knows? Maybe the UK, I think they did do some changes now because I think London really lost out on being an IPO venue over the past five years. Yeah. And so, but still, I think there's so much activity and capital in the US that it's likely going to be NASDAQ.

27:52Florian Faes:All right.

27:53Esther Bond:That would be great for everybody. and would be a very big vote of confidence in the industry. So LTPs, come on, come on. Go to Goldman, go to Bank of America. List your company now, please. All right. What's your final prediction?

28:12Florian Faes:Yeah, I feel like this one's less punchy, but I'm just going to say continued consolidation of the market generally in the form of acquisition of smaller boutique LSIs by, I'm going to say specifically by mid-sized competitors. I mean, maybe this doesn't, maybe again, this is sort of obvious, but I'm talking sort of, you know, translation agencies, for example, in the region of two to five million, probably being acquired by a larger competitor, maybe more so tech enabled in the region of 10 to 20 million. I think we're going to see consolidation happening at those levels.

28:52Esther Bond:So if you are a smaller LSI and you do want to sell, who do you call? Who do you email? Email esther at slater.com. Great. All right. So call to action. If that's what you're planning, email esther at slater.com and we can help you through that journey. So final prediction, which is a little bit out of my technical depth, but I thought I threw it in there given all the macro stuff we've talked about that XLift2 may replace or at least strongly compete with TMX in just general localization management. There's an interesting YouTube by the Blackbird people, Blackbird team on on YouTube about this topic, they're pushing XLIF2, which is, you know, kind of translation data management or exchange format, right?

29:49Esther Bond:So, they're saying that's the TLDR of the podcast, that TMX is outdated, lacks standard maintenance. XLIF2 is kind of seen as the next generation standard for translation memory exchange. Positives is it prevents loss of data by avoiding lossy conversions required by TMX, has superior data structure for translation tools for LTPs, apparently separates inline text and handles tags much more clearly. And crucially, the format is much better suited for AI training because it has richer contextual info, so a lot of more metadata in the XLift file that would arguably make it much more suitable for contextual translation, which is what we're already all looking at.

30:35Esther Bond:So, XLift2 may replace TMX.

30:38Florian Faes:I think I remember when TMX was a new thing. That's kind of scary. Now it's outdated. Although, you know, things happen quickly. Yeah, technology becomes legacy rapidly. Cool.

30:51Esther Bond:All right. So, you're going to have to live with this podcast for like three weeks, dear listeners and viewers, because, you know, we all need a break and we're going to go on a two-week break over the holidays and we'll be back in January. Not sure if we managed to record a podcast in the first week, but definitely in the second week we'll be back online. So, yeah. Thanks for this, Esther. And to everybody, have a great rest of the year and, you know, Happy New Year.

From the publisher

In the 2025 year-end episode of SlatorPod, hosts Florian Faes and Esther Bond reflect on a year defined by rapid AI investment, shifting policy, and structural change across the language industry.

Esther opens the year-in-review by highlighting January’s twin funding milestones in the language AI and product space. Florian follows with February, which saw hyperscalers and AI labs release data highly relevant to the way AI translation is being used.

March, April, and May saw major developments both on the regulatory side and in terms of bolt-on acquisition deals.

Past the mid-year point, OpenAI’s decision to hire a localization manager was what grabbed the industry’s collective attention. The AI lab’s decision contrasted with September’s news, which saw the closure of one of the world’s most recognized academic programs for localization.

The year closed on publicly listed LSIs releasing mixed results and major announcements in AI translation for literature and live speech translation rollouts.

The duo closes with 2026 predictions!

More from SlatorPod

All 39 episodes
2025 Recap and 2026 Predictions!SlatorPod · 31 min
Listen in VO