# Anthropic CEO Dario Amodei & Accenture CEO Julie Sweet Talk New AI Deal | Bloomberg Talks

*Bloomberg Podcasts* · https://www.youtube.com/watch?v=ZC7CzeOqlwc

## Summary

Bloomberg interviews Accenture CEO Julie Sweet and Anthropic CEO Dario Amodei about their new multi-year partnership to accelerate enterprise AI adoption. The deal creates an Accenture Anthropic business group that will train around 30,000 professionals to use Claude and help clients show AI returns in the P&L. Amodei also discusses donating the MCP protocol to a foundation, and declines to confirm reports of an early IPO.

## Key points

- Accenture and Anthropic announced a multi-year partnership forming the Accenture Anthropic business group to drive enterprise AI adoption.
- The partnership aims to train around 30,000 professionals to use Claude and manage the full lifecycle of enterprise deployment, tracking ROI with CIOs.
- Amodei says Anthropic's revenue grew from a 1 billion annualized rate at the start of the year to 10 billion at the end, with 80% of that being enterprise revenue.
- Julie Sweet says Accenture has trained 500,000 people in AI fundamentals since 2019 and has begun agentic AI training for its roughly 750,000 'reinventors'.
- Sweet argues it is too early to call the end of entry-level jobs, saying Accenture is adding communication training so new hires can operate at a higher level.
- Anthropic is donating MCP, the model context protocol, along with other things to a foundation to standardize how AI models connect with enterprise data.
- Amodei declined to confirm reports of early IPO talks, saying the company is focused on growing enterprise revenue.

## Notes

### The Deal
- Accenture and Anthropic announced a multi-year partnership to drive enterprise AI adoption.
- They will form the Accenture Anthropic business group and train around 30,000 professionals to leverage AI.
- Julie Sweet says all partnerships start with client demand, and the goal is helping clients make the "P&L pivot" — showing AI results in the P&L.
- The deal coexists with Accenture's expanded OpenAI partnership because clients expect Accenture to be the number one partner for every technology company that matters.

### Anthropic's Enterprise Push
- Dario Amodei says adoption will start on the coding side, citing Claude Code and large productivity gains.
- The aim is to expand Claude usage into healthcare, life sciences, financial services, and public sector.
- Anthropic will work with CIOs to track deployment and returns, managing the whole lifecycle with Accenture.
- Amodei notes Anthropic's revenue went from a 1 billion annualized rate at the beginning of the year to 10 billion at the end, with 80% enterprise revenue.
- He estimates hundreds of billions, maybe trillions, of enterprise productivity value remain to be created, which is why partnering with Accenture matters.

### Talent and Entry-Level Jobs
- Sweet says Accenture is focused on creating sustainable entry-level jobs, calling new hires "reinventors."
- Since AI can let people do in months what took two years, Accenture is adding communication training to close the experience gap.
- She says it is too early to call the end of entry-level jobs; companies must change how they train.
- Accenture introduced classical AI training in 2019, had 500,000 people trained in AI fundamentals before ChatGPT, and has kicked off agentic AI training for its roughly 750,000 people.

### MCP and Standardization
- Anthropic is donating MCP, the model context protocol, along with other things to a foundation, alongside others including Google and OpenAI.
- Amodei compares an AI model to a smart brain that still needs an integration layer to access enterprise knowledge — MCP standardizes how models intersect with data.

### IPO
- Asked about reports of early IPO talks for 2026, Amodei said the company is currently focused on growing the business and enterprise revenue.

## Transcript

[00:00] [music] Bloomberg Audio Studios, podcasts, radio, news. >> Accenture and Anthropic inking a multi-year partnership deal to drive enterprise AI adoption. The two, they're going to form the Accenture Anthropic business group with the goal of training around 30,000 professionals in leveraging AI as a catalyst for reinvention. pleased to say joining us right now is the CEO of Accenture, Julia Julie, and the CEO of Anthropic, Daario Amade. Great to have both of you here. Uh, and Julie, I'll start with you. Uh, mainly because when I saw this, I thought, "Wow, this is neat." Then I also thought about the fact that just last week you announced a further expansion of your partnership with OpenAI. Talk a little bit about how

[00:46] these two coexists under the Accenture umbrella. >> Sure. So, um, all of our partnerships start with client demand. So the partnerships are responding to client demand and we have a long history. Our clients expect us to be the number one partner for every technology company that matters and uh Enthropic this partnership that we've announced today is uh very important to our clients. And so that's the role we play is to make sure that we can take the world's best technology and mix it up with industry knowledge with process and how to actually use it in the enterprise to get to an outcome. And as you know today it's all about when is AI showing up in the P&L. We call that the P&L pivot and this partnership uh will help our

[01:33] clients make that pivot. Well, I am curious then Dario if you can kind of talk about specifically the anthropic tools, the idea here uh of standardizing how some of these AI agents actually connect to uh the various uh uh tentacles if you will, the various applications out there. >> Um yes, so we are um uh you know I think we're going to start on the coding side. Some of the technologies that have been adopted quickly have been uh uh things like uh claude code the the the um tool tool we produce to help with development and we've seen very large productivity gains with that but we want to go beyond that to more general use of our tool uh claude and we want to drive usage of the tool across areas like healthcare um

[02:15] life sciences financial services public sector you know things that anthropic has worked with for a long time you know more broadly I would say that uh enterprises have been experimenting with AI for quite a while, but uh you know I think really you know as Julie said with the with the P&L the the important thing is really to deploy these at scale and you know to actually drive ROI for them. And so that's what we're trying to do. That's why we have this whole business center. That's why we're training 30,000 people to use Claude. And we're going to work with CIOS of companies to track how how how they're using uh Claude through Accenture, how they're uh deploying the tools, what returns they're getting,

[02:55] what returns they're getting from the tools. So, we're going to work with Accenture to manage the whole life cycle. We think these tools are are are are revolutionary. We've seen that from their use inside Anthropic and many enterprises already. [snorts] But, uh, the thing we're most excited for is is to make sure that, you know, we can we can really, uh, uh, spread these across the enterprise world. And, you know, that's that's where Accenture comes in and can really drive these things at a scale that hasn't been done before. >> Yeah, it's interesting, Dario. I mean, you think about AI, some of these uh, LLMs, they're very popular with consumers, individuals. Uh, they've been well socialized. But when it comes to

[03:32] enterprises, when it comes to that business customer, you touched on maybe some of the skepticisms that CIOS have in really adopting AI as you look into 2026, this deal with Accenture. Uh how much further does it put you when it comes to sort of unlocking the enterprise world and what's the goal when it comes to 2026? >> Yes. So, um you know, we we've already seen substantial enterprise adoption. This year, Anthropics revenue went from 1 billion uh annualized rate at the beginning of the year to 10 billion annualized rate at the end of the year. Um, so we've already seen a lot of adoption and you know 80 80% of that revenue is enterprise and within that particularly with things like coding but

[04:13] in other areas we're we're starting to see very strong ROI but the scale of this is so large right the scale of what needs to be done right there's hundreds of billions maybe trillions of of you know kind of enterprise enterprise productivity out there value to be created and so that while while it has been growing quickly and we've been overcoming that skepticism I think what needs to be done h has you know is enormous in scale compared to what to what has been done what has been done already and that's why partnerships with Accenture are so important that you know we really need to get to a scale that it's it's you know a young company needs to needs to partner with someone who really understands the enterprise world

[04:53] >> and Julie you made the point that you know one of the big questions here is when we're going to see AI show up uh in the P&L there's also the big question of when you see AI really show up in the job market and I'm curious you know whether you're seeing AI slow down hiring for entry- levelvel jobs. Over at Accenture, for example, [snorts] >> we are really focused on creating sustainable entry-level jobs with AI. And so our pivot in the talent side is how do we now take our entrylevel uh uh uh people, we call them reinventors coming in and give them the skills so they can operate at a higher level. So for example, if you can now do things that maybe would have taken you two years, uh then you're losing two

[05:37] years of experience at client. So how do we now intro we're now introducing communication training to help close that gap? And so we believe it's way too early to call the end of entrylevel jobs versus the responsibility that we have as a company and as companies across the globe to change the way that we train and that we think about jobs. so that we have sustainable entry level jobs. >> I am curious though uh Julie as you sort of prep your own uh folks in internally for uh not only using the tools yourselves but obviously the idea that these tools ultimately will then somehow be used by clients. What has been the uptake internally? How much retraining have you had to do for your own folks? >> Our training has been massive. Uh it

[06:27] starts with back in 2019 we introduced uh classical AI training. So when chatpt came on that we were already had 500,000 people trained in the fundamentals of AI in addition to the specialists. We just kicked off our entire 750,000 or so reinventors agentic AI training and in fact this partnership is all about tapping into the talent advantage that Accenture brings because you need great technology expertise you also need the industry expertise like in banking you need to know your be able to know your customer and so we're training all of our people those who understand industry and function and that's a big talent advantage. So Dario, but this goes back to the idea kind of of what you're contributing and the idea that a

[07:13] lot of people have been looking for more of a standardized model to be able to work off of and obviously anthropic at least the way as you guys promote yourselves. The idea of open source, the idea of maybe a little bit more transparency. There was a report today in your information that you along with a couple of others uh including uh Google and open AAI for that matter are going to be part of this AI foundation that will seek to create a more standardized system for some of this. Can you talk a little bit about that? >> Yes. So, so this is um we're we're donating uh MCP which is the called the model context protocol along with some other things to a to a to a foundation. And and what MCP is, it's it's a tool

[07:55] that uh uh uh kind of governs and standardizes how AI models intersect with data. If you think about it, you can think of an AI model as kind of as kind of the brain, right? You have a smart brain that knows a lot of things. And when I bring that brain into an enterprise context, there's all this kind of knowledge and and tacit knowledge that is present in the enterprise. And so bringing them together is always a challenge. You know, the way I often like to say it is, you know, suppose I took some very smart person off the street and, you know, plopped them into the middle of my enterprise, gave them an executive role and say, "Okay, congratulations. You know, you're in you're in charge. You get to make all the decisions." Now, as

[08:32] smart as that person is, they wouldn't at least not right away make make make make make immediately good immediately good decisions. So, there's this integration layer that's missing and and you know, I think these tools are designed to facilitate that integration layer and and make it easier. And so we're we're excited to provide these things to the world and and you know kind of kind of make it easier to facilitate access to the technology across the world. >> And Dario, before we let the two of you go, I have to ask while we have you, there has been reporting that you've been in early talks for an IPO as early as 2026. Can you confirm that? Uh currently we are we are just focused on growing our business um uh and you know

[09:11] growing growing growing enterprise uh growing enterprise revenue as as you can see in this partnership. >> All right. Well, obviously it's a great partnership. I've known Julie for some time and certainly uh you picked a good partner there. Dario Amade is the CEO of Anthropic. Julia, great to see you once again. and Julie, the CEO over at Accenture.