HomeVideos

US Orders Anthropic to Disable Foreign Access to Mythos | Bloomberg Businessweek

Now Playing

US Orders Anthropic to Disable Foreign Access to Mythos | Bloomberg Businessweek

Transcript

1151 segments

0:02

Bloomberg Audio Studios, podcasts,

0:05

radio, news.

0:08

This is Bloomberg Business Week Daily.

0:11

Reporting from the magazine that helps

0:13

global leaders stay ahead with insight

0:16

on the people, companies, and trends

0:18

shaping today's complex economy. Plus,

0:21

global business, finance, and tech news

0:23

as it happens. The Bloomberg Business

0:26

Week Daily podcast with Carol Masser and

0:29

Tim Stenbec on Bloomberg Radio.

0:32

>> Just a warning to everybody. We're going

0:33

to get a little technical, but Maggie's

0:35

going to help us understand what

0:36

jailbreaking is and what's at stake when

0:39

it comes to anthropic, Fable 5, and

0:41

Mythos. Anthropic executives planning to

0:44

meet with Trump administration officials

0:45

on Monday to discuss an unprecedented US

0:47

government directive barring foreign

0:49

access to the company's most advanced AI

0:51

models. That's according to a person

0:53

familiar with the matter. For more,

0:54

we're joined by Maggie Eland, Bloomberg

0:56

News tech reporter. She's based in

0:57

Washington DC. It's where she joins us

0:59

from right now. So, Maggie, I said this

1:01

would get a little technical because I

1:03

think we should just start off by

1:04

explaining the Fable 5's relationship to

1:07

to mythos. Uh, and what has the

1:10

government right now so concerned and

1:12

why anthropic is pushing back on this?

1:15

>> Yeah, absolutely. It definitely helps to

1:17

sort of get some definitions down to

1:19

start. So, Mythos that was released in

1:21

April by Anthropic. It's their latest

1:23

model, but it was not released to the

1:25

public because the company was too

1:27

worried that it had a lot of cyber

1:29

capabilities and it could be used for

1:31

cyber attacks. So, it was only released

1:34

to a limited group of government and

1:36

business partners to sort of help those

1:38

partners shore up their own cyber

1:40

security efforts. What happened last

1:42

week was Anthropic actually released a

1:45

version of that model with some added

1:47

safeguards to the public. So that's what

1:49

we call fable 5 and that's sort of what

1:52

started this whole situation between the

1:55

government and anthropic. So a few days

1:57

after that public release, the

1:59

government sent an enthropic a letter

2:02

that was an export control directive

2:04

that essentially blocked anyone who was

2:06

a foreign national from even accessing

2:09

that fable 5 or that mythos 5 model. So

2:13

in effect, Anthropic was forced to

2:16

immediately disable those those models

2:19

uh especially since they have foreign

2:21

nationals who are employees. So this

2:22

affects their employees, it affects

2:24

their customers. All of a sudden they're

2:26

only able to provide these models to US

2:28

citizens and they don't even have a way

2:31

currently of sort of a know your

2:33

customer standard in terms of you know

2:35

where those people reside who are using

2:37

their models.

2:38

>> I'm glad you said that because I'm I'm

2:39

opening an Claude uh from Anthropic

2:41

right now on my phone. And I I don't pay

2:43

for it, but I but I have it. I can use

2:44

it. I I I I had the ability to use Fable

2:48

5 if I paid for it, but now it says

2:50

currently unavailable. It also

2:52

advertises that it's for your your

2:53

toughest challenges. But Claude has no

2:55

idea where I live in terms of like what

2:58

my nationality is. So it they just have

3:00

to disable it because they don't have

3:01

those know your customer guard rails in

3:04

place.

3:06

>> Exactly. Um it would definitely take

3:08

some time to develop those. there could

3:10

be some privacy concerns from folks in

3:13

Washington. It certainly wasn't

3:14

something that they could do on very

3:17

short notice. Um, and they were they

3:19

were sent this letter on at a pretty

3:21

quick turnaround. Um, so, you know, as

3:24

we've reported, anthropic executives are

3:26

now in Washington DC. They're meeting

3:29

with the Commerce Department today.

3:31

We'll see what comes of that. They're

3:33

they're trying to work through this. The

3:35

concern here, I should point out, and

3:36

Anthropic has said this in their

3:38

statement. They believe that the the

3:40

government is concerned about a

3:42

jailbreak. Now, there's not necessarily

3:44

one definition for what that is, but

3:46

typically it means that users who are

3:49

using Fable might be able to prompt it

3:52

in a certain way so that they're able to

3:54

access those cyber advanced capabilities

3:56

that they're not supposed to access. So

3:59

Microsoft was involved in surfacing this

4:02

vulnerability and that's certainly going

4:04

to be dominating the conversation uh in

4:07

terms of just trying to figure out if

4:09

this is something Enthropic could fix,

4:11

if they could potentially fix it with

4:13

more of these know your customer

4:14

standards. Uh so that's what's going on

4:16

right now and and we'll have to see if

4:18

if they're able to to work this out and

4:20

somehow remove that export control. So

4:22

on one hand, you know, Maggie, it seems

4:25

like the White House, the Trump

4:27

administration

4:28

is actually very keen to what could

4:31

possibly happen and is staying ahead of

4:33

the game. It almost sounds like um like

4:38

who is leading kind of the strategy and

4:41

the process when it comes to the White

4:43

House? And like it sounds like the

4:46

expertise is there. Is that correct?

4:48

because I feel like with social media,

4:49

we felt like maybe everybody didn't

4:51

quite understand everything that was

4:52

going on. Um, and so there's been some

4:55

consequences as a result, but it does

4:57

sound like the team at the White House

4:59

and the administration is kind of on top

5:01

of things. Is that fair?

5:04

>> Yeah, I think uh a few things like at

5:06

the principal level after Mythos was

5:08

released in April, you did kind of see

5:10

this full government response. You know,

5:13

Susie Wilds, Trump's chief of staff got

5:15

involved. Scott Besson, the the Treasury

5:18

Secretary, Sean Korncro, the the

5:20

director of the national cyber um

5:24

efforts. So, you saw a lot of actors

5:26

coming together very quickly to

5:28

essentially write an executive order to

5:32

make sure that the government could have

5:33

early access to these cyber capable

5:36

models, including Mythos. So, this cyber

5:39

use case kind of like scared everyone,

5:42

woke everyone up uh across the Trump

5:44

administration. Um uh but what we're

5:46

seeing now is a little bit more

5:48

confusion I guess and and it's unclear

5:50

exactly what the policy is because that

5:52

executive order in the process of

5:54

writing it there was some internal

5:56

policy debate over whether this should

5:58

be a mandatory regime or a voluntary

6:00

one. And ultimately what you saw in that

6:03

executive order in early June was a

6:05

completely voluntary uh situation where

6:07

AI model companies on you know on their

6:10

own terms would offer their models to

6:12

the government and the government would

6:14

sort of review them help them ensure

6:16

that everything was safe. At the same

6:18

time the government would be able to

6:20

make sure that critical infrastructure

6:22

and its own data was protected from

6:24

potential cyber attacks. And this was

6:26

all going to be sort of a kumbaya

6:28

voluntary thing. And I guess the lack of

6:30

clarity now is was what we're seeing is

6:33

it's really no longer voluntary for

6:34

Anthropic in this case. Um when push

6:37

came to shove, the company was was doing

6:39

something that the government didn't

6:41

like uh and then ultimately found a tool

6:44

to sort of compel it into paying more

6:47

attention to this jailbreak. So that's

6:49

definitely casting some uncertainty onto

6:52

the whole industry. Uh because Anthropic

6:54

is not the only one sort of operating

6:56

under this voluntary situation. And it

6:58

seems that potentially it's it's more uh

7:01

voluntary but with a looming threat of

7:03

you know you could be compelled to do

7:05

something if if the government decides

7:08

to do that.

7:09

>> Maggie I I'm curious the folks you talk

7:11

to who are observers of this type of

7:14

technology and understand tech policy.

7:17

Do any of them say that that we're

7:18

missing the mark here that because what

7:20

what I think about is the the constant

7:23

refrain which is like the wor you're

7:24

using right now the worst version of AI

7:27

that that will be available. You have

7:29

all these companies here in the US that

7:30

are working on this. You have all these

7:31

companies in China that are working on

7:33

this and here we are talking about some

7:34

model that a year from now is going to

7:36

look like pretty basic and simple if if

7:39

the last three years are you know any

7:41

indication of how quickly this

7:42

technology moves. Are we just missing

7:43

the mark here by focusing on this one

7:45

thing?

7:46

>> Yeah.

7:47

Yeah, I mean I think there's been a lot

7:50

of of criticism both like in the cyber

7:52

security community and in the AI policy

7:55

community to this decision. Um I think

7:58

the main criticism is is just that it's

8:00

a bit incoherent, right? Sort of like

8:03

this is probably the worst thing for AI

8:05

model companies is like regulatory

8:07

uncertainty because this was done

8:09

through essentially a letter so it's not

8:12

public. Um and andropic has said that it

8:15

wasn't even clear exactly why this was

8:17

done. They they only found out later um

8:19

that is their position. They found out

8:20

later that uh the government was worried

8:23

about this vulnerability. So certainly

8:25

some folks have said that the Trump

8:27

administration is missing the mark

8:28

because you know if they wanted to do a

8:31

more regulatory approach uh they should

8:33

make that clear right because that would

8:34

be the best thing for the markets is to

8:37

have um just a clear signal in terms of

8:39

what the regulation was going to be. um

8:41

or if they were going to take sort of

8:43

the pro-innovation lzair approach then

8:45

sort of doing something like this in a

8:47

lastm minute export control blocking all

8:50

foreign access is certainly not aligned

8:52

with that. So I think the main criticism

8:54

right now is just uh the lack of

8:57

coherence which is probably the the

8:59

worst thing for the market even compared

9:01

to a more clear uh regulatory policy.

9:05

>> Yeah. Yeah. And it does wonder, you

9:06

know, as you guys you and the team, um,

9:09

in your reporting, Maggie, say, after

9:10

decades of disseminating cutting edge US

9:12

technology around the world as a

9:14

diplomatic and economic tool, Washington

9:16

is now moving in the opposite direction.

9:18

And you do wonder uh is it not just a

9:21

Trump administration shift but maybe

9:23

something that on both sides of the

9:25

political aisle uh there's been a

9:27

pulling back because it's safe to say

9:28

there was certainly you know the

9:30

continuation of um Trump's first term

9:33

policies when it came to China on

9:35

certain things and technology in the

9:37

Biden administration. So you do wonder

9:39

if this is something more significant

9:40

going on.

9:42

>> Yeah, there there's definitely this

9:43

unique tension with artificial

9:45

intelligence in particular. um where

9:48

folks in Washington are concerned that

9:51

actually, you know, disseminating this

9:53

technology might be sort of at odds uh

9:56

with actually preventing foreign actors

9:58

like China and and Russia from from

10:00

accessing it. Um you know, the US

10:02

adversaries, the people that they don't

10:04

want to access it. So, this is kind of a

10:06

unique tension. Um, and it's been, you

10:10

know, currently as we're seeing, it's

10:11

it's sort of leaning in the direction of

10:13

of safety, which is certainly not what

10:15

we would have expected from the Trump

10:16

administration um, at the start of the

10:18

president's term. Uh, but I think it's,

10:21

you know, it's one thing to say that you

10:22

want to have, you know, a light touch

10:25

approach. It's another to actually be

10:27

sort of confronted with these cyber

10:29

vulnerabilities, um, as folks like Scott

10:32

Besson have been and then to sort of

10:34

figure it out on the fly,

10:36

>> right? And you have to wonder about

10:37

especially as a lot of these companies

10:38

are getting ready to do some maybe mega

10:40

cap um IPOs. What does it mean for their

10:43

business going forward? Uh some great

10:45

reporting. Maggie, thank you so much. We

10:47

really appreciate it. Maggie Eastland,

10:48

she's Bloomberg News tech reporter.

10:50

She's based in the nation's capital, our

10:52

Bloomberg News Bureau in Washington DC.

10:55

>> Stay with us. More from Bloomberg

10:56

Business Week Daily coming up after

10:58

this.

11:01

>> You're listening to the Bloomberg

11:03

Business Week Daily podcast. Catch us

11:06

live weekday afternoons from 2:00 to

11:08

5:00 Eastern.

11:08

>> Listen on Apple CarPlay and Android Auto

11:11

with the Bloomberg Business App

11:12

>> or watch us live on YouTube.

11:16

Um, it's interesting. We're obviously

11:17

watching the war and we just kind of

11:19

covered all of that in terms of the

11:20

market enthusiasm that's playing out.

11:22

Also playing out is again we're focusing

11:26

on what is the world's largest IPO.

11:27

We're talking about SpaceX successful

11:29

launch on Friday. We continue to see the

11:31

stock shooting higher today. shares

11:33

jumping in their second day of trading,

11:35

adding to gains following a blockbuster

11:37

debut that instantly vaulted this

11:39

company into the ranks of the world's

11:40

most valuable public companies. Now,

11:42

keep in mind the offering syndicate for

11:44

the IPO included several lead

11:46

underwriters including JP Morgan Chase.

11:48

The list to be fair though of all the

11:50

banks involved reads like a who's who of

11:52

Wall Street's biggest and best known

11:54

banks.

11:55

>> We got a great couple of voices joining

11:56

us with an inside look at how all of

11:58

this played out and continues to play

12:00

out at this point. All right, Billy Lip

12:01

Schulz is Bloomberg News IPO reporter.

12:03

David Bower joins us too, head of Equity

12:05

Capital Markets America's over at JP

12:07

Morgan Chase. They both join us here in

12:08

the Bloomberg Business Week studio. Uh

12:11

Dave, good to see you. Uh Bailey, I feel

12:13

like we

12:14

>> usual. Yeah, it's like I badge in, I

12:17

always seeing you and we're lucky. Uh

12:19

SpaceX, of course, the largest IPO ever.

12:22

I'm just curious from from your view how

12:24

you think market participants right now,

12:25

Dave, should be looking at at this in

12:28

the context not just of other potential

12:30

IPOs in the future, but really in the

12:33

context of everything we're seeing out

12:35

there right now, the talk of people

12:37

taking money from other assets, selling

12:39

that, getting into SpaceX. Just your

12:41

view.

12:41

>> Yeah, I think I mean, first of all, for

12:44

somebody in the ECM world, what a time

12:46

to be alive. Like these are these are

12:47

great times for the markets. And I think

12:49

unlike other cycles we've seen where IPO

12:51

issuance has been extremely active, you

12:54

have a real catalyst forming the

12:56

investment thesis behind this. This is

12:57

the reindustrialization of America.

12:59

We're creating new ecosystems, new

13:01

economies. I mean, you think about what

13:03

space could become, what AI could

13:05

become. These are these are investment

13:08

thesises that many people haven't been

13:10

able to invest behind before. And so, I

13:11

think seeing that this euphoria in the

13:13

IPO market is warranted and appropriate.

13:16

What's the signal to the market and

13:17

market participants specifically? I

13:19

>> I I think right now you're you're seeing

13:21

that the IPO worked and you saw Friday a

13:23

very strong trading with the stock, you

13:24

know, finishing up 19% from the IPO

13:26

price. And I think today you see, you

13:28

know, broader markets rallying on the on

13:30

the news with um uh with, you know, Iran

13:33

peace treaty and you're seeing just more

13:35

enthusiasm coming in. So I think

13:36

>> SpaceX up another 15% today.

13:38

>> Exactly. And so I think it's all it's

13:39

all systems go to make a you know a pun

13:41

intended for SpaceX. But I think if

13:42

there was any uh hesitancy of you know

13:45

should I continue to buy up 20% I think

13:47

the answer is yes. And if you look long

13:48

enough there's a you know this is a

13:51

business that could be generationally

13:52

transformational.

13:53

>> But thinking about the generational

13:54

transformation Dave when I look at the

13:56

pipeline or the kind of group of

13:58

companies that could be coming public

14:00

they don't look like SpaceX. When you're

14:03

meeting with people maybe at your former

14:04

employer KKR talking about something in

14:06

their portfolio, a number of other

14:08

private equity firms, what are they

14:10

looking at? Obviously, this is a big one

14:11

for IPO buyers, but does that translate

14:14

to someone who owns say software

14:16

companies or other companies in other

14:18

industries? Look, look, the the mega

14:20

IPOs are getting the headlines right

14:22

now, but there's been a very active and

14:25

very uh accommodative uh equity market

14:27

in general throughout this year. And

14:29

putting SpaceX aside,

14:31

uh, new issuance volumes are up almost

14:33

2x year this time from last year. And

14:36

the vast propoundonderance of that has

14:37

been other sectors and we've seen

14:39

healthcare coming back, we've seen

14:41

biotech issuance coming, we're seeing

14:43

uh, industrial, energy, all so the

14:46

markets are working at large.

14:47

>> Are all those themes still off of AI in

14:50

some capacity? If the bottleneck is

14:51

power and I'm a power company, then I at

14:53

least have an AI pitch because when I

14:55

look at air talk to folks, it's kind of

14:57

like your TAM is either infinite because

14:59

you have AI behind it or it's zero

15:01

because your software and the worry is

15:03

that your kind of total market could be

15:06

at risk. How do you think about kind of

15:07

other industries and how that fits into

15:09

that?

15:09

>> Yeah, I look I I don't think the market

15:11

is shut out for certain issuers. Um, I

15:13

think it does come down to price and

15:16

having the right starting point in the

15:17

public markets and I think there is a

15:18

bid for those other businesses. Um, you

15:21

know, you mentioned sponsor back

15:23

businesses. Um, it's a good time to, you

15:25

know, what I would say get the puck on

15:26

the ice. Start it. And, you know, we

15:28

always say when we're advising our

15:30

clients and our issuers,

15:32

>> proof points in the public market are

15:34

worth even more than in the private

15:35

market. And so, getting out there,

15:37

starting it, and you can see how, you

15:39

know, your valuation can expand you uh,

15:41

as you perform. um you can get your

15:43

share price to a level that you might be

15:44

more interested in selling at, but

15:46

getting, you know, getting started can

15:47

be helpful.

15:48

>> What is it about this moment right now,

15:49

Dave, that is compelling so many of

15:51

these companies not just to go public,

15:52

but also to look into going public? And

15:54

I'm I am going back to sort of the the

15:56

companies that when they do go public

15:58

will become mega cap companies.

16:00

>> And just to tag on to what Tim's saying,

16:02

opportunistic or fundamentally based? I

16:04

>> I think I think this is a more

16:06

fundamental based market. And you you

16:08

look back to 2021 which felt a little

16:11

bit more technical like rates were zero

16:13

and so therefore equities were

16:14

attractive and you had a very different

16:16

dynamic that drove that. I I think this

16:18

is much more of a fundamental. I think

16:19

the market is looking at what the next 3

16:22

to 5 years could be and you know

16:24

potentially looking past some of the

16:26

near-term volatility and it's not saying

16:27

they're ignoring downside risk but I

16:29

think saying there's certain businesses

16:31

right now that are very much worth um

16:33

investing in. There's a whole slate of

16:35

new issuance that is different than the

16:37

portfolio I've I've had in the past and

16:38

I'm going to take that opportunity to

16:39

invest in it.

16:40

>> I am thinking though about those mega

16:41

IPOs be it anthropic or open AI. Should

16:44

we make any assumptions about their

16:46

market reception just because of what

16:48

happened with SpaceX?

16:50

>> I I think I mean SpaceX is is is an end

16:53

of one of itself um from the fact that

16:55

no one else is doing space at this scale

16:57

at that this velocity. Um

16:59

>> but it's an AI play

17:00

>> but it is an AI play. where I was going

17:02

with that is I do think I mean this

17:04

bodess if you're anthropic and open AI

17:06

you're you're applauding that this type

17:07

of market reception happened and you

17:09

know the the size of the deal getting

17:10

done having a trade as well as it did uh

17:13

to me that that gives even more

17:14

confidence that the next wave uh could

17:16

get done in a very uh positive way.

17:18

Well, when we look at what the next few

17:20

months can look like, back to the

17:22

broadening out, is AI still the flavor

17:24

of the day? If we look at some of the

17:25

reports that are out there and SKHix or

17:27

other large companies in the AI space

17:30

looking to tap the market like Carol was

17:32

me mentioning

17:33

>> because of these tailwinds and kind of

17:35

what does that mean for this July, this

17:37

August class?

17:38

>> Yeah, I look I I think we're going to

17:39

have a very active summer. Um, even even

17:42

the SpaceX period of time, you know, a

17:43

lot of people thought coming in we'd

17:45

have a der of issuance and you'd have a

17:46

real quiet period. We did $10 billion of

17:49

equity capital outside of SpaceX last

17:51

week. And so, you know, the market was

17:53

still working. There were still a ton of

17:54

deals getting done. Um, I think this

17:56

capex cycle to your point of of AI,

17:59

people still want to invest on it. And

18:00

that's across equity, debt, um, you

18:03

know, all the facets of capital here and

18:04

so the public markets are are helping

18:06

that, but it's also the private capital

18:08

markets as well. You're we're seeing

18:09

capital formation in almost every corner

18:11

of of the capital markets.

18:12

>> You oversee ECM. I come back to the

18:14

point that Google, Alphabet, however you

18:16

want to call them, raised close to $90

18:18

billion across a suite of products. Is

18:20

that something we should expect from all

18:22

these hyperscalers who need capital the

18:24

t the equity market and is there any

18:26

risk that that gets oversaturated?

18:29

>> I think that the the speed at which

18:32

Google is able to raise in the public

18:33

markets shows the depth um and the

18:36

capacity, you know, for for the capital

18:37

markets. And I think as long as

18:39

companies are showing an ROI and a good

18:41

use of proceeds to raise that equity

18:43

capital, the markets will be there for

18:44

them.

18:45

>> How do we know when it becomes

18:49

uncomfortable, crazy, exuberant,

18:52

>> frothy,

18:52

>> frothy?

18:53

>> Well, think back to 20 and 21. What it's

18:55

easy now to say that these were

18:56

companies that were advantageous, but in

18:58

the moment, did it feel that way? I I go

19:01

back to when you look at fundamentals

19:03

and you look at valuations,

19:05

we're not terribly stretched and we're

19:07

not at a, you know, a a new point in

19:09

time. And you go back and you look at,

19:10

you know, for example, software

19:12

valuations in 2021, those were hitting

19:14

new highs and they were hitting kind of

19:15

new records of where the market was

19:16

trading at. We're not seeing that at

19:18

large in the public market.

19:20

>> Train isn't profitable.

19:21

>> No, it's not. But it's and it's being

19:22

valued on a sales multiple for 2025

19:25

sales and and that and I think that some

19:27

people have come and sat here and said,

19:28

you know, it's starting to feel like,

19:30

you know, when we're valuing IPOs on

19:31

sales multiples, it started to feel a

19:33

little bit like the dot boom. Why is

19:36

this time different?

19:37

>> I I well, one I think I I think that's a

19:40

unique one. SpaceX, I think, is its own

19:41

animal. And I think when you're thinking

19:42

about the space economy, you're you have

19:44

to look at that in a very different way.

19:46

I don't think you're seeing the rest of

19:47

the market being valued aggressively on

19:49

a sales m multiple or other metrics

19:51

where people are trying to extend and

19:53

get comfortable with it. I think these

19:54

are based more in fundamentals. It's

19:56

based on growth right now and people are

19:58

seeing where the capital being laid

20:00

today has a return in the future and I

20:02

think you can pull that forward and

20:03

that's what's going on.

20:04

>> But the circular financing doesn't worry

20:06

you guys or what are the what what's the

20:08

we have this conversation a lot. Are we

20:10

stupid to have that conversation about a

20:13

company that seems to buy from another

20:14

part of its properties or invests in a

20:16

chipmaker because they need to like you

20:18

know what I'm saying?

20:20

>> What's the conversation you guys have

20:22

about that circular financing?

20:24

>> Uh the the what keeps me sleeping well

20:26

at night about this is the fact that

20:27

we're having the conversation. And I

20:29

think if if people are acknowledging it,

20:30

we're talking about it, you're

20:32

dissecting it, you're diligencing it, I

20:34

think the cycle has room to continue and

20:36

and to grow. I think it's the unknown

20:38

risks, not less the the known risks that

20:41

um I think you know could could derail

20:43

the cycle.

20:44

>> And when we look at this market, is

20:47

there a risk and and kind of just

20:48

thinking about the difference between a

20:50

hyperscaler who is historically been the

20:52

best cash flow cows in the history of

20:54

mankind, they can at least turn off

20:56

spending. How do you think about

20:58

companies that are spending and need to

21:00

spend but don't have hundreds of

21:03

billions of dollars annually or by

21:04

annually in terms of free cash flow?

21:06

>> Yeah, I look I think you um you have to

21:09

look at the the the the fundamentals of

21:11

the contracts. What what is what is it

21:13

contract you know what is contracted

21:14

from the demand build that they have

21:16

today and get comfortable that what

21:18

they're investing in right now has the

21:19

right economics to yield a return in

21:21

three or five years. Um and I think

21:24

you're right like you know they might

21:25

not have the uh free cash flow spet

21:27

today but you look at um what is

21:29

contracted and what could be p you know

21:31

come to fruition in a high quality way

21:33

in the next two to three years um you

21:35

you can bridge to that free cash flow.

21:37

>> We got 30 seconds left here. You've been

21:39

covering this market. What do you want

21:40

to ask Dave?

21:41

>> Do we see more private equity IPOs in

21:43

the second half or is that still an area

21:45

broadly speaking that's touch and go?

21:47

>> I I I think definitively yes. Um I I

21:50

think the market um would be

21:52

accommodative uh for that and and our

21:54

backlog suggests that we will have uh a

21:56

number of those in the second half

21:57

>> in about 10 seconds. How much does debt

21:59

leverage or how much does leverage

22:00

ratios matter for those companies?

22:02

>> Uh it it matters I think um I think you

22:04

want to get it to a comfortable starting

22:06

point but I think for good free cash

22:08

flow stories with solid predictable

22:10

revenue and growth the market can get

22:12

comfortable.

22:13

>> So welcome to our regular weekly

22:14

segment. We'll see you both here back on

22:16

Monday. belly lip shelves of course

22:18

Bloomberg News IPO reporter Dave Bower

22:20

head of equity capital markets America's

22:22

over at JP Morgan Chase.

22:24

>> Stay with us more from Bloomberg

22:26

Business Week Daily coming up after

22:28

this.

22:32

>> You're listening to the Bloomberg

22:34

Business Week Daily podcast. Catch us

22:36

live weekday afternoons from 2:00 to

22:38

5:00 Eastern. Listen on Apple CarPlay

22:40

and Android Auto with the Bloomberg

22:42

Business App or watch us live on

22:44

YouTube.

22:46

Hey, speaking of sports, maybe that's

22:48

one of the reasons that Fox is buying

22:49

Roku at a $22 billion value in this push

22:53

for streaming video. As we just

22:55

mentioned, shares of Fox fell 17% on

22:58

this news. Shares of Roku for much of

23:00

the day were little changed. They ended

23:03

up closing today, just down by about 2%

23:06

on the day today, but a lot of this was

23:07

priced in earlier this week. Keep in

23:09

mind Roku on Friday when it started to

23:11

leak out on Friday um when we were on

23:14

air uh that stock was up 20% in the

23:16

trade. So um a lot of the news was

23:19

already I think it's safe to say

23:20

factored in.

23:21

>> I want to bring in Githa Ranganathan

23:22

Bloomberg Intelligence senior media

23:23

analyst. She joins us from Princeton,

23:25

New Jersey. Githa, what is Fox going to

23:28

do with Roku?

23:31

>> So uh Tim, they're promising that

23:33

they're going to keep Roku as is. Uh so

23:36

really what Fox is buying here is

23:38

distribution. They own the content. They

23:40

own some of the best live content in

23:43

terms of access to sports properties, in

23:45

terms of their uh news coverage with the

23:47

Fox News Channel. But what they really

23:50

don't have is that distribution network

23:52

and that is really hurting them because

23:54

they have huge exposure to the linear TV

23:57

ecosystem. Uh where really nobody is

23:59

really staying there anymore. everybody

24:01

has kind of migrated away from, you

24:03

know, the linear TV ecosystem and

24:04

continues to migrate migrate away from

24:06

there to to streaming. And so they want

24:08

to go where the action is and they want

24:09

to go where the money is and where the

24:11

eyeballs are. And that's really what

24:12

they're getting with Roku. Uh and so far

24:15

they're promising that they're going to

24:16

keep operating it just as it is right

24:18

now.

24:19

>> What about Fox one? What about 2B?

24:22

>> Again, we'll continue to operate as is.

24:25

Uh but you know, again, we we don't know

24:28

how things change. Obviously, they're

24:29

saying things that need to be said uh

24:32

because Roku does partner with everybody

24:34

in the ecosystem. And I think some

24:36

people are worried that maybe, you know,

24:38

Roku's always kind of been known as this

24:40

neutral platform, right? Uh

24:41

>> Switzerland, the Switzerland of the

24:43

streaming box, that was something they

24:45

sold to like me as a consumer tech

24:46

reporter like 15 years ago, you know, as

24:49

as Chrome was or as Google was coming

24:51

out with the chcast, as Amazon was

24:53

coming out with its streaming stick,

24:54

they were like, "Hey, we got no skin in

24:56

this game. We work equally well with

24:58

everybody.

25:00

>> Yeah, exactly. And so, uh, you know,

25:02

this is this is a little bit of a

25:04

headscratcher in terms of how all of

25:06

that neutrality is going to work. Uh,

25:08

but, you know, from what Fox Management

25:10

has said so far, uh, they said they're

25:12

going to keep operating the platform as

25:14

it is. Of course, we have to see how,

25:17

you know, the other content companies

25:19

feel about it. I mean, this is kind of

25:20

letting the the fox into the hen house,

25:22

if you will, pardon the pun, but that's

25:24

really what it feels like a little bit,

25:26

but we'll have to we'll have to see.

25:28

>> Um,

25:30

so does a deal like this mean somebody

25:32

else I always like wonder about that.

25:33

Does anybody else have to do something

25:34

or no? This is really Fox playing

25:36

catch-up to some extent.

25:38

>> So Fox Yes. Yes. Um, so Fox uh stayed

25:42

away from all of those expensive

25:44

streaming wars when everybody was going

25:45

crazy and launching, you know, a plus

25:48

channel, Disney Plus, Paramount Plus,

25:50

you know, uh, Peacock or all all of the

25:52

streaming platforms. Fox was the one

25:53

that kind of very famously stayed away

25:56

from all of that and said, and actually

25:59

in in in hindsight, that was the smart

26:01

move to make because all of those

26:03

companies lost billions uh if not tens

26:06

of billions of dollars on making those

26:08

very very expensive foray into

26:10

streaming. Um, and it all ultimately

26:12

came down to profits. And so at the end,

26:14

Fox actually ended up looking looking

26:16

like a winner in many ways because they

26:17

hadn't lost that much of money. But I

26:19

think the one thing that has kind of uh

26:21

definitely weighed on investors mind and

26:23

generally weighed on sentiment has been

26:25

that Fox has really heavy exposure to

26:27

the linear TV ecosystem. One that you

26:29

know they they really needed to kind of

26:31

reset their narrative. Um and the Roku

26:34

acquisition helps them do that. So right

26:35

now they get 90% of their revenue from

26:38

linear TV only about 10% from digital.

26:41

This move immediately helps them get

26:44

close to about 35% of their revenue from

26:46

from digital. But more importantly, it

26:49

really helps them kind of be there at

26:51

the center of it all, right? They

26:52

they're in on all types of action with

26:54

streaming, whether it's advertising and

26:56

whether it's subscriptions. So, anybody

26:58

who gets a subscription through a a Roku

27:01

uh box, you know, Fox is going to be

27:03

able to participate in those economics

27:04

and and that matters.

27:05

>> That was my family on Saturday night

27:07

when we wanted to watch this final game

27:10

of the Knicks and Spurs. We spent 30 or

27:13

35 bucks on ESPN and we bought it

27:15

through Roku. So yes, I don't know who

27:17

who knows. Um I I I I wonder what this

27:21

mean I mean what what's so this is

27:23

company like in the early days of

27:25

streaming Anthony Wood is the CEO the

27:27

founder of the company. He he like

27:29

invented the way to get streaming to the

27:31

biggest TV in your home. This guy's like

27:32

legendary when it comes to streaming.

27:34

This started out as like an experiment

27:36

at Netflix to try to get Netflix on on

27:38

the biggest screen. Are there any

27:40

antitrust issues here because it is such

27:42

a big gatekeeper?

27:44

Yeah, they are a huge gatekeeper and one

27:46

can argue that yes, this with this deal

27:48

that you know Fox does kind of become

27:50

this vertically integrated platform.

27:52

Remember they will get access to 100

27:54

million global streaming households and

27:56

you know we're talking almost close to

27:58

about 70 million you know US broadband

28:00

households. That's a pretty sizable

28:02

number. That's half of you know

28:03

basically US broadband households. But

28:05

you know the one thing that that we have

28:07

to keep in mind is the Murdochs are

28:09

really cozy with you know the current

28:11

administration with uh the President

28:13

Trump and you know in this environment

28:16

we've we've kind of seen those uh

28:19

friendships actually go a long way. So

28:21

maybe they get uh they they don't go

28:24

through so much of scrutiny but

28:25

obviously there is always that that that

28:27

chance that there might be some

28:29

regulatory uh watch as well.

28:31

>> Was Roku worth 22 billion? Mhm.

28:35

We actually think it might be worth more

28:38

um and and I say that because you know

28:41

um it just if you just look at the

28:44

growth profile of this company Carol

28:46

it's been amazing. So the way they have

28:48

been really honing in on some of their

28:51

undermonetized assets. So they've

28:53

refreshed their home screen and this is

28:55

the first time that they're doing this.

28:56

This is their biggest product refresh in

28:58

about 10 years. And what it really helps

29:01

them do is kind of become like this

29:03

Netflix landing page. So you know that

29:05

the Netflix landing page is very much

29:07

like how you can make or break careers,

29:10

right? You can make or break a show. You

29:11

can, you know, cotton just pops up there

29:13

and that's it. Everybody's talking about

29:16

that. That's that's the subject of all

29:17

the water cooler conversations. And so

29:19

Roku is very much now trying to be that

29:22

Netflix homepage. So you can really see

29:24

them uh they've rolled out the new

29:26

homepage to about 20% of their of their

29:28

base. uh it it's going to be rolled out

29:31

to to to the rest of their users. Uh but

29:34

it's all really about elevating the

29:36

content, right? And um and people are

29:38

willing to pay for this prime, you know,

29:41

inventory. Um and we're seeing that

29:44

actually in the numbers. So if you just

29:46

look at Fox or any traditional media

29:48

company, we're kind of, you know, if you

29:50

get flattish revenue growth, that's a

29:51

big deal. Roku's platform revenue is

29:53

expected to grow about 21% this year.

29:55

21%. Yeah. So that is that is a big

29:58

deal. That's impressive. Um, Gita, thank

30:00

you so much. Gita Rangan, she is

30:02

Bloomberg Intelligence senior media

30:04

analyst joining us uh from our BI

30:06

offices in Princeton, New Jersey.

30:09

>> Stay with us. More from Bloomberg

30:11

Business Week Daily coming up after

30:12

this.

30:17

>> You're listening to the Bloomberg

30:19

Business Week Daily podcast. Catch us

30:21

live weekday afternoons from 2:00 to

30:23

5:00 Eastern. Listen on Apple CarPlay

30:25

and Android Auto with the Bloomberg

30:27

Business App or watch us live on

30:29

YouTube.

30:31

>> Carol, for for so long,

30:33

>> we've had people come into this studio,

30:35

the

30:35

>> also the other iteration of our

30:37

Bloomberg Interactive Broker Studio

30:38

where Charlie is.

30:40

>> Yes.

30:40

>> Holding down the fort. And one of their

30:43

main propositions was these were wealth

30:45

managers and one of their their main

30:46

things was like there fewer and fewer

30:48

companies are going public. There's less

30:50

public stock out there because companies

30:52

are buying back shares. We are

30:54

increasingly getting into alternatives

30:57

because there are fewer public companies

30:59

out there,

31:00

>> right? Because we need to put money to

31:02

work to invest

31:04

>> and that's been one of the most defining

31:05

mo like elements of the equity market

31:07

over the past couple of decades. So

31:10

writes Carmen Reinicki Bloomberg new

31:11

stocks reporter. She's among the authors

31:14

of today's big take. It is one of the

31:16

most read stories on the Bloomberg

31:17

terminal. She joins us here in the

31:18

Bloomberg Interactive Brokers Studio.

31:20

Suddenly,

31:22

there's a lot of value. Well, I don't

31:25

want to use the word value because I

31:26

don't want to get people, you know,

31:27

thinking value versus growth, but

31:29

there's a a lot out there for investors

31:31

to buy. What's what did your analysis

31:33

that you did along with Lu Wang find?

31:35

>> So, basically, there's been sort of a

31:37

switch in like expansion and contraction

31:40

of the market. So, over the last few

31:42

decades, we've seen companies really

31:44

aggressively buy back shares, right? So

31:46

sort of shrinking the overall pool of

31:48

stocks available and giving a lot of

31:49

support to investors along the way. This

31:51

year however with all of these new IPOs

31:54

that are coming down the pipe. I mean we

31:55

just saw record from SpaceX. Um we'll

31:59

hopefully see Anthropic and OpenAI by

32:01

the end of the year and we've seen

32:04

equity issuances from companies like

32:05

Alphabet. um the you know the number of

32:09

stocks or things available to trade in

32:11

the market is expanding to the tune I

32:12

think it's 1.5 trillion of stock added

32:16

to the US equity market over the next

32:18

two years that's from JP Morgan Chase

32:20

and Co.

32:21

>> Is this just cyclical though where you

32:23

know at at a certain point when when

32:25

these companies see value in buying back

32:27

their shares they're going to they're

32:28

going to gobble at them up again.

32:30

>> I think there are a lot of factors that

32:32

kind of go into it. One of it is yeah a

32:34

lot of these companies especially in the

32:35

technology space have these huge balance

32:37

sheets. There were you know not other

32:39

places I guess that they were wanting to

32:41

deploy capital. So buying back shares

32:44

was a great way to return some of that

32:45

that value to shareholders. Now they

32:48

have these huge spending needs because

32:50

of AI you know building out the data

32:52

centers and all the other technology

32:54

that they need to use it. And so they

32:56

want to be spending money and raising

32:59

money to spend on those things. So the

33:01

part of it is that and then I think on

33:03

the flip side, you know, this is

33:04

something that's sort of similar to what

33:06

we saw in the 1990s sort of around the

33:08

like the dot boom and bubble bursting.

33:12

So I think there is some cyclicality

33:14

there as well, although people are

33:16

obviously very very careful to, you

33:18

know, necessarily compare them too much

33:20

or say that this time isn't different if

33:21

we're talking about a bubble.

33:22

>> Is there any worry um that there won't

33:25

be enough investors to buy all all the

33:27

stock that's issued?

33:28

>> That's definitely a concern. So, I mean,

33:31

we had a really great sort of run out of

33:33

the gate with SpaceX, which definitely

33:35

assuaged some fears that there wouldn't

33:37

be enough investor demand or that the

33:39

market would have trouble sort of

33:41

digesting this large of an issue.

33:43

>> I was shocked how well it all went down

33:45

on Friday. I was shocked.

33:47

>> I was too. I mean, I think everyone was

33:48

sort of braced for a lot more issues and

33:50

it went very smoothly and even today we

33:52

had a second great day of trading.

33:54

There's clearly a lot of demand there.

33:56

Um but as we get through the summer, the

33:59

next couple of months of the year, uh

34:01

the lock lock lockups are going to start

34:03

to expire. So a lot more of SpaceX stock

34:05

is going to come into the market. Its

34:07

float right now is quite low. And at the

34:09

same time, you're going to be getting,

34:10

you know, these other IPOs, Enthropic,

34:13

OpenAI, and it I mean, we could see

34:16

other companies also issuing secondary

34:18

offerings or see just you know, this

34:20

pool of in the equity market grow even

34:22

more. So that is I think a concern

34:24

hanging over the market for the you know

34:26

the rest of the year probably if not

34:27

more.

34:28

>> I'm glad you just reminded everybody

34:29

that it's only the second day of trading

34:30

for for SpaceX because I mean even even

34:32

a company like Cerebrus right we you

34:34

know how much did we talk about this

34:36

company in the first few days of trading

34:37

and it did hit a high on May 14th so

34:42

exactly a month ago $311 a share. It's

34:44

down to $218 a share

34:46

>> right now. And and I don't know what the

34:48

lockup structure is and who was able to

34:49

sell and when they were able to sell but

34:51

you're right. I mean, two days of

34:54

trading for a new company doesn't

34:55

necessarily make the the history.

34:58

>> No, it doesn't. And it it's even hard to

35:00

necessarily call it a trend. Uh, so

35:02

obviously it's a stock that we're

35:04

watching very closely over the next few

35:06

days, weeks, months, years.

35:08

>> One one thing that I want to talk about

35:09

in your piece is this idea of companies

35:11

staying private for longer. And and

35:14

again, years we heard that companies

35:17

would stay private for longer because

35:18

for a few reasons. one uh quarterly

35:20

earnings. We love them as reporters and

35:23

getting to talk about this stuff,

35:24

>> but it makes it tricky for companies to

35:26

grow their business, right?

35:28

>> Yeah. I mean, they're, you know, they're

35:30

marking a market every second of every

35:32

trading day. It's it's not an easy thing

35:34

to have happen and it's great for

35:35

liquidity and if they want to raise

35:37

money, but often times you can raise a

35:39

lot of money in the private market and

35:40

stay look I mean SpaceX is from 2002. A

35:43

company is

35:44

>> old. That's an older company than

35:45

Facebook,

35:46

>> right? Exactly.

35:47

>> Yeah. So, it showed that it could be

35:48

private for a long period of time. Does

35:50

this get companies off the sidelines?

35:52

You mentioned Anthropic and and um Open

35:54

AI, but other than those,

35:55

>> I think we'll see. So, something I've

35:57

heard from a lot of investors is that

35:58

SpaceX kind of opened the IPO window. Um

36:01

we've definitely seen, you know, fewer

36:03

than normal, I think, IPOs since the

36:05

Spack craze of 2021. And so this

36:07

hopefully kicks off a period of a lot

36:09

more companies coming to the public

36:11

markets, not just these huge mega cap

36:13

technology companies, but um one sort of

36:15

across you know the wide range of

36:17

sectors and sizes um available in the

36:20

market and and that would be a really

36:21

good thing for investors. The other

36:23

thing that I I hear from sources is that

36:25

they worry a little bit about all of

36:27

this value being created in the private

36:29

market that I mean it really leaves the

36:31

public market out and in some cases some

36:34

argue that it can inflate these

36:36

valuations without sort of giving

36:38

everyone a chance to be weighing in

36:40

which is you know basically what the the

36:42

public stock market is doing. It's also

36:44

a reminder, you know, of how expensive

36:47

it is for this build, right, in terms of

36:49

AI. And it also is a reminder that when

36:51

you really need a ton of cash, right,

36:54

the capital markets or the or the public

36:55

market is really where you go

36:58

ultimately.

36:59

>> Yeah. And equ offering equity like doing

37:02

these offerings has become cheaper in

37:04

some aspects than debt. So, it makes

37:08

sense that these companies would be

37:09

looking to to go to the equity markets.

37:12

And I mean they're doing both, right?

37:14

But it makes sense that they'd be

37:16

specifically wanting to tap equity

37:18

markets.

37:18

>> I just think it's fascinating how

37:19

>> secondaries. Is that what you're going

37:20

to say?

37:21

>> Secondaries, but also just how it's

37:23

debt, it's equity. Like you're just

37:25

seeing

37:26

>> the all the throw everything at the wall

37:27

approach of raising capital.

37:29

>> Just reaching out to every bucket that's

37:31

there in terms of raising money and

37:33

capital for this build. And I guess it's

37:35

just something we'll watch really

37:36

carefully because you you look at um was

37:39

it the Nvidia, right?

37:41

today, right? And three times overs

37:43

subscribed like that.

37:44

>> And that was a debt raise.

37:46

>> That was a debt raise, but it's just you

37:48

see the investor interest

37:50

>> and and we certainly saw it with SpaceX

37:52

and we'll see what else. We're just

37:53

essentially an AI play, right?

37:55

>> It is and it isn't. This is something

37:57

I've been asking people a lot about as

37:59

well. And I think there's a little bit

38:01

of tension out there. I mean, certainly

38:03

the market seems to be valuing it as an

38:04

AI company as much as a space company,

38:06

but I I've definitely heard some sources

38:08

say, you know, their their AI offering

38:11

is so much behind the others.

38:14

>> Yeah, it's got to deliver right before

38:16

Kerman.

38:18

This is the Bloomberg Business Week

38:20

Daily podcast available on Apple,

38:23

Spotify, and anywhere else you get your

38:25

podcasts. Listen live weekday afternoons

38:28

from 2 to 5:00 p p.m. Eastern on

38:30

Bloomberg.com, the iheartradio app,

38:33

TuneIn, and the Bloomberg business app.

38:36

You can also watch us live every weekday

38:38

on YouTube and always on the Bloomberg

38:41

terminal.

Interactive Summary

This episode of Bloomberg Business Week Daily covers the significant regulatory tensions between the Trump administration and AI company Anthropic, concerning export controls on advanced models due to cybersecurity fears. Additionally, the podcast explores the current surge in IPO activity, highlighted by SpaceX's successful market debut, and analyzes how traditional media companies like Fox are attempting to pivot by acquiring digital platforms like Roku to remain competitive in a landscape shifting toward streaming.

Suggested questions

3 ready-made prompts