00:00 Speaker A
Microsoft is teaming up with Anthropic to integrate Claude Co-work into its 365 co-pilot platform. The partnership aims to improve task efficiency across Microsoft software, such as building Excel spreadsheets and scheduling meetings using AI. Dan Halle spoke to Microsoft’s president of Business apps and agents about this and what this means for the tech giant’s AI strategy. Take a listen.
00:32 Charles Lamanna
What we’ve done there is you’ve really taken co-pilot from being primarily a chat-based experience, you know, asking questions, generating emails. I mean all about actions. So, sending emails, generating documents, updating your schedule and your calendar. and we did that in close partnership with Anthropic.
00:54 Dan Halle
And so, how does this kind of work? You know, we we’ve seen, uh, how co-work and Anthropic can kind of do things for you on its own, you know, the automated tasks. How does that work then with with Microsoft 365? What what kind of tasks can you expect it to kind of complete for you?
01:21 Charles Lamanna
Yeah, a big thing about Copilot co-work is that it integrates with all the M365 applications and all of the data that you have stored inside of office apps. So we call that work IQ. So it’s really seamlessly integrated with all of those office applications and workflows we know and love. Plus it’s hosted in the cloud. So you don’t have to run it locally on your device. You can just run it, close your laptop screen, walk somewhere else and it’ll keep churning away to complete the task for you.
01:51 Dan Halle
And so I guess, you know, if I’m getting a bunch of emails or I have to set up a meeting with someone, I need a presentation. I’d seen a demo of this uh that you guys have put out. It it’s basically, you know, I just tell Copilot co-work, hey, put this, you know, presentation together, get these people uh together for a meeting later this afternoon, and it’ll just go ahead and do that for me while I can go grab a coffee or something.
02:22 Charles Lamanna
That’s exactly right. And uh I think if you can think of it and you could probably dispatch it to another person, you probably can give it to co-work and the thing that we’ve really seen as people have used it the last couple of weeks is you can dispatch a lot of these things at the same time. So, it’s not like generate one presentation or generate one spreadsheet or schedule one meeting, it’s more like one after another and you can let five, six, seven of these things run around in the background.
02:49 Dan Halle
What what are you seeing as far as uptake from enterprises when it comes to these different AI offerings? You know, Microsoft has Copilot, you know, built into to 365. Uh, you know, it it allows you to work throughout the different suites, uh, the the the suite itself, whether that’s, you know, with word or or Outlook. What what’s the uptake been though for for enterprise users?
03:20 Charles Lamanna
Yeah, I think we’ve been super impressed. Just in the last year, we’ve seen a 10 X increase in the daily active users for M365 Copilot. 90% of the Fortune 500 are using M365 Copilot. So I think we’re in a very rapid adoption phase and diffusion of these tools at work.
03:45 Dan Halle
And you know, I think a lot of people when they think Microsoft and AI, they, you know, they think open AI. You guys have that that big deal together. You’ve been working uh side side side, uh, you know, hand in hand for for some time now. Now we’re we’re talking anthropic. I guess, what does that mean for for Microsoft? Does it mean, you know, a pivot to more models? What what does that look like?
04:13 Charles Lamanna
Yeah, Our big focus is we’re going to be multi-model and we’re going to use the right AI model for the job. So, for example, inside of M365 Copilot, we have an auto router, it’ll choose, in some cases it may use an open AI model, other cases it may use an anthropic model. Um, and the real magic comes when you take those models and you join it together with your office data, that work IQ that we have.
04:47 Dan Halle
And is that, I guess, you know, when when you go in yourself, it does that auto routing, can you select whether you want open AI or or if you want Anthropic models when you just, you know, kind of navigating on your own?
05:03 Charles Lamanna
Yeah, and end users can go pick, we have like a drop down, you can go choose, of course, but we try to keep auto because the best model for the job may change on a day-to-day basis or a month-to month basis. Like the best way to chat or the best way to create a spreadsheet may be one model in March, it may be a different model in April.
05:22 Dan Halle
When when it comes to these kinds of, you know, developments and apps, everything obviously is going incredibly fast. Uh, I think that’s the one thing everyone can agree on when it comes to AI is that it continues to evolve quite rapidly. Where do you see this going now for for Microsoft when it comes to these kinds of agents, when it comes to these capabilities and and the productivity side of things?
06:04 Charles Lamanna
Yeah, we think that agents are going to show up everywhere you work because another big part of the announcements today is we have agent capabilities inside Excel, inside Word, inside PowerPoint. So you can ask questions, but it can also start updating formulas or changing your tables, create a pivot table for you or update your Word document. So we think this idea of what you see with co-pilot co-work is going to show up in lots of different places, dedicated and embedded and it’s what people are going to start to expect from AI tools, just like you expect chat to work.
06:33 Dan Halle
So it’s going to be essentially, you know, when you think Microsoft 365, you think for people like me office, uh I grew up with with with office. You think about that that suite of tools, eventually, do you just say, okay, well AI is a part of it, right? Like, you know, just as word or Excel is a part of it, AI is now a part of it and that’s, you know, it’s part and parcel for it. Is that how it kind of goes eventually?
07:07 Charles Lamanna
That’s exactly right. I mean, you have Word, you have Excel, you have PowerPoint, you have co-pilot. Like that’s what modern office and information work is going to look like.
07:18 Speaker C
Dan Halle, with me now to unpack this great interview. Dan, here’s a question. So, let me start simple. I’m a viewer. I just listened to this very good interview. Great interview. Great interview. Maybe I’m an office worker. What what does this tool do for me? Like how does it make my life easier?
07:46 Dan Halle
It’s supposed to be basically automate a lot of the tasks that you would do throughout the day that maybe you don’t necessarily have time for or are just a pain. Right? So, basically the way that they’d explained it uh in one of their uh releases was that you would be able to say, you know, tell it that you need to uh prepare a presentation for, you know, uh a uh meeting and then send out uh emails inviting folks in your team to that meeting uh and it would then go off and do that. And it’s not as though it just does one thing, it could do multiple things at the same time. So, you know, you can tell it, oh go do this and then you know what, go do that too and then and then this. Now, of course, you know, a lot of this is stuff that uh you know, Anthropic has announced with with Claude before um and co-work. Uh and you know, I I think part of this feeds into that uh or or reduces that narrative that software was, you know, done for, right? That it was basically, you know, going the way to dinosaurs that, you know, and the uh enhances that idea that instead, AI is just going to be kind of uh an additive to them, right? And you know, we’ve talked about that how, you know, we saw the software stocks take huge hits when Anthropic announced co-work. Uh what was it? Uh uh in January and then they announced the the plugins for it and that just sent people into a tizzy. Uh and you know, service now got crushed, uh salesforce got crushed uh into it uh times of Reuters. But, you know, since then I think people have come to realize, oh well, it’s just going to go into this stuff and they’ll they’ll sign deals with other companies. And yeah, they may have their own offerings down the road, but it makes more sense for it to work in something like, you know, uh Microsoft 365 which, look, I don’t know why Microsoft, this is just a tangent. I don’t know why Microsoft decided to change the name from Office. You had the name. That was it. Everybody knew Office. It doesn’t need to see be Microsoft 365. Just call it Office, guys. It’s fine. Everybody’s using it in the office anyway. Uh and so that’s the that’s the the general idea that you’ll just throw that into there and you know, you’ll have Microsoft 365/ Office doing all of this work and have AI along the way. And to me that made sense from the getgo, not that, you know, Anthropic would come in and say, all right, not only are we working on, you know, these AI models, but we’re also just going to build from scratch this uh you know, uh productivity suite. and oh yeah, a customer relationship management system and things like that. So, you know, this seems to be the way that the the the industry is more likely to go.
09:33 Speaker C
Why did Microsoft feel the need to team up with Anthropic? They couldn’t use their own AI models.
09:39 Dan Halle
Yeah, so they’re they have their own AI, but I I think they see that Anthropic is more advanced. Uh it’s one of the most advanced labs out there. And then, oh yeah, they have that huge deal with Open AI, which has kind of been like, you know, they’re like neighbors, but like, it feels like the neighbor that you let borrow your lawn mower and then you get it back and you’re like, he scuffed it over here. Man, I hate that guy. It’s it’s like they’re they’re not exactly best friends uh anymore by it feels by any stretch of the imagination. Now they’ll probably tell you, you know, nah, we’re we we love working together. But it’s, you know, it seems to have become acrimonious as as time’s gone on. But, you know, they still have access to Open AI’s capabilities. Um Open AI is kind of spreading its wings a little more. Uh and so Microsoft is basically hedging and saying, look, we can’t just rely on this one company. You know, uh we we should be offering these different models uh to to customers and, you know, as Charles was saying, allow them to kind of, you know, automatically uh go between, you know, Open AI or or or Anthropic or, you know, what have you whenever the the the application sees which one is better for that particular task. And so, you know, it is interesting that they have Anthropic just because it’s, you know, obviously they don’t like each other, Open AI and Anthropic. Uh and Microsoft and Open AI have had this kind of weird, you know, relationship now. Uh and they’re now working with Anthropic big time. So it’s it’s fun to watch this kind of little soap opera among these, you know, massive companies.
11:15 Speaker C
A final question, kind of a bigger question. I mean investors have this concern. Some do. I mean, the stock is down like 15% here this year. All this AI spend where the ROI is. Do the co-pilot stats, which you you talked about in your article, Dan. So 15 million paid seats, 160% quarter over quarter growth. Does that in your opinion help address that?
11:39 Dan Halle
I think that it does to a degree, right? I mean, the I think that investors are going to want to continue to see improvement over time, obviously, that’s just, you know, how investors are. Uh you know, the the idea that this is going to help increase growth for Microsoft’s cloud business, that’s what all of this is about, right? And so, you know, and it’s it’s almost it came at kind of the perfect time for their cloud business just because we’d seen that the growth slow, right? Because, you know, you hit a critical mass, massive numbers, and so naturally you’re just going to start slowing as far as growth goes. This though, another shot in the arm, right? It’s another amount of money that companies have to pay each, you know, per user, per month. And so it it helps amplify that that growth and and uh uh uh kind of re-trigger that. I think uh what they need to show is that kind of similar trajectory that they had with cloud. Otherwise, I mean, they’re spending billions and billions of dollars, right? I mean, you know, they have commitments but that’s from Open AI. Uh people kind of look at that and say, well, okay, so where’s everyone else on there, right? And so they’re also in this three-way race with Amazon and Google to be the AI kind of, you know, hyper scale leader, right? And so, at this point, Google is back in the lead. Um we had been saying that Open AI was, you know, crushing them. And then Google came out with Gemini 3 and it’s kind of been in Google’s, you know, court since then. Um And then you have Anthropic coming in and just saying, you know, okay, here’s our stuff and everybody freaking out. So, I don’t think anybody really knows what to make of of AI at this point, but they just want to see these companies continue to grow revenue.
13:30 Speaker C
Dan, great interview, great article. Thank you, my friend.
