Return to Transcripts main page


Interview With "The Ezra Klein Show" Host And The New York Times Columnist Ezra Klein; Interview With George Washington University Professor Of Law Catherine J. Ross; Interview With The Washington Post Reporter Shane Harris. Aired 1-2p ET

Aired April 14, 2023 - 13:00:00   ET




BIANNA GOLODRYGA, CNN SENIOR GLOBAL AFFAIRS ANALYST: Hello, everyone and welcome to AMANPOUR. Here's what's coming up.



GUTHRIE: Months of protests come to a head as France rules on its polarizing pension reform. We're on the ground in Paris.

Then, is the race to unlock artificial intelligence spiraling out of control. I asked "New York Times" journalist and top podcast host Ezra


Also ahead, as the billion-dollar lawsuit against Fox News gets underway, could first amendment rights be one of the big losers? I speak to freedom

of speech expert Catherine J. Ross.



UNIDENTIFIED MALE: This was a deliberate criminal.


GOLODRYGA: The 21-year-old accused of leaking classified U.S. intel makes his first appearance in court. Walter Isaacson gets the details with Shane

Harris, one of the journalists who first broke this story.

Welcome to the program everyone. I'm Bianna Golodryga in New York, sitting in for Christiane Amanpour.

Well, France's highest constitutional court has just approved a pension reform plan that is triggered months of anger and discontent. The

controversial bill which will raise the retirement from 62 to 64 could be enacted as soon as this weekend. Now, the new law has put French President

Emmanuel Macron under immense pressure, but he says it is essential. Demonstrators are already back on the streets and voicing their opposition.

Correspondent Fred Pleitgen is on the ground for us in Paris. Fred so, what is the response been behind you to this ruling?

FREDERIK PLEITGEN, CNN SENIOR INTERNATIONAL CORRESPONDENT: Well, yes, Bianna, it's been one of extreme anger from the folks who have come out

here. I'm actually in the square in front of city hall. And if you look around, you can see that the square is pretty full now. There's a lot of

people with signs. I would say the folks that we're seeing out here, they really come from all walks of life and they are of all ages, which is, of

course, something that's very important to point out when you're speaking about pension reform. It is also a lot of younger people who are coming out

on the streets here, student groups trade unions.

And if you look at some of the signs, you can see some of the people here at city hall, they've climbed up on this sign for the Olympics. There's a

sign there that says, climat de colere, climate of anger. Behind that, you can see two smaller signs, one that says, c'est la fin du chemin

democratique, that means, this is the end of the democratic path. And then the one next to that says, la vraie democratie elle est ici, and shows

basically, the people.

So, essentially what the people here are saying is that they are not going to give up fighting against this pension reform. A lot of the folks that we

have been speaking to had said, Bianna, that they had already expected that the pension reform was going to be OK'ed by the constitutional council, but

they don't see that as legitimate. They believe that the constitutional council, essentially, supports the politicians in this country, supports

the government of Emmanuel Macron.

Of course, we know that some of the folks who are on that council are political appointees by Emmanuel Macron. So, as you can see, it is a lot of

anger that is being unleashed here by a lot of people who are on the ground, and it's something that we've seen over the past couple of months.

And while some of the numbers, Bianna, had been dwindling, I was at the protest, big protest that took place yesterday where 390,000 people went on

the streets all over France.

You can still see that people are coming out here in full force. And the message that they are sending tonight to the folks of the constitutional

council and, of course, to the president of this country as well is that this is not over. They are going to continue to fight this reform bill. Of

course, the bill itself makes a lot of people very angry, but then also the way that it was pushed through by Emmanuel Macron, using some of those

executive powers and essentially bypassing large parts of the legislative process. Bianna.

GOLODRYGA: It's a bigger picture, a win, no doubt, for Emmanuel Macron, raising the retirement age to 64 from 62. But it is notable that the

council struck down six measures that were tied to reforms that they found to fall outside the scope of the law by France's constitution today as

well. So, I'm wondering what the impact of that is amongst those crowds behind you because it is --

PLEITGEN: Yes, that's --

GOLODRYGA: -- it is worth noting that they're more subdued today and less angry, it appears, than they were in weeks prior.

PLEITGEN: Yes, I mean that -- that's something that some people expected as well. That they thought that some of those measures that were put inside

that reform bill would be shot down by the constitutional council because they were deemed to not be part of a reform bill that would raise the

retirement age or in general about the retirement age.


Certainly, that is something that has uplifted some people. But you're absolutely right. The people are -- I wouldn't necessarily say less

subdued, they certainly are very loud. The protests here is very colorful, it's not certainly as violent as some of the things that we've seen over

the past couple of months, and certainly some of the things that we saw yesterday.

However, it is still also very early here today. Nevertheless, I do have to say that there are still a lot of people who believe here on the street

that they can still make a big difference, that this law can still be struck down. There's one gentleman that I spoke to who said that, look,

there have been measures in the past that have been pushed through and that because there was so much anger. So many protests coming that in the end,

those laws were repealed or the government took them back. They believe that that is something that could happen.

And one of the things, I think, that we also need to point out, Bianna, that I think is very important is that there are, of course, also divisions

here among politicians as well. One of the things that was very interesting is when it was announced that the constitutional council had OK'ed this


There were two signs that were unrolled here on Paris city hall, and they say, mairie solidaire avec le mouvement social, saying that the mayor is in

accordance and supports the social movement, which is obviously the social movement that you see right here. And, of course, the mayor of Paris, a

very powerful politician here in this country, Anne Hidalgo. She has, from the very beginning, said that she is against this pension reform.

However, there is the flip side of that as well. There are a lot of people who say they believe because Emmanuel Macron and his government have been

so badly damaged politically, by the way that this reform bill has been pushed through, that the far right are the ones who are going to benefit

from this. And we've already heard from Marine Le Pen, the leader of the far right here in this country, saying essentially, look, people need to

vote for her instead of people like Emmanuel Macron in the future.

GOLODRYGA: Yes. Well, one can surmise that for opponents and proponents and advocates of this reform. Nonetheless, it is a monumental day and shift

in that country. Fred Pleitgen, thank you.

Well, next to an emerging technology that could leave humans with a little more R and R or potentially end of life as we know it. That's one way to

put it. The race to develop artificial intelligence and the more advanced artificial general intelligence is moving at a frantic pace. Faster than we

or even its creators can really understand and grasp. Unsurprisingly there are fears about what this tech means for the future. It's an unease that's

deeply embedded in human culture and express famously in films like "The Matrix" and "2001: A Space Odyssey". Remember this?


UNIDENTIFIED MALE: Do you read me, Hal?

DOUGLAS RAIN, ACTOR, "2001: A SPACE ODYSSEY": Affirmative, Dave. I read you.

KEIR DULLEA, ACTOR, "2001: A SPACE ODYSSEY": Open the pod bay doors, Hal.

RAIN: I'm sorry, Dave. I'm afraid I can't do that.

DULLEA: What's the problem?

RAIN: I think you know what the problem is, just as well as I do.


GOLODRYGA: Well, now, we separate fact from fiction with "New York Times" columnist Ezra Klein. He's spoken to experts on all sides of this debate on

his popular podcast, "The Ezra Klein Show". And he joins me now from San Francisco.

It's really good to see you, Ezra. So, first I'm wondering if you can just explain for our viewers how A.I. is already impacting their daily lives.

EZRA KLEIN, HOST, "THE EZRA KLEIN SHOW" AND COLUMNIST, THE NEW YORK TIMES: So, the problem with A.I. is it's a very broad term that describes a lot of

different ending. So, one thing you have in there is machine learning, right? Fundamentally what we're dealing with are algorithms that are

trained on some kind of data set and they learned how to predict something. The next token in a sequence, the next word in a sentence, the next image,

or they know how to predict some kind of outcome of an equation.

So, machine learning is used in all kinds of things. When you run around the internet, and every ad seems to know what you clicked on before, that

is all machine learning, in a way that is A.I. The thing that has broken this whole conversation open and it led to people feeling like the future

might be very, very different than the past is a rise of what are called large language models.

So, these are A.I.s that are trained on truly massive quantities of usually digital text. They have an amount of compute power behind them that boggles

the mind. And what's happening there is they are learning relationships between -- I mean, what are to them statistical concepts in a way that is

allowing them to communicate pretty naturally with human beings and prove able to complete a very, very wide variety of tasks so long as you can

basically do them on a computer.

And it is the general nature of these coming systems and the rapidity of their improvement that has shifted the conversation from OK, we got

computer said, you know, run equations to, are we dealing with some kind of new intelligence or at least some kind of transformational technological


GOLODRYGA: So, is this what I referenced earlier is artificial general intelligence?


And if so, how do you broadly explained the difference from A.I. as we know it right now and what many expect to see in the future?

KLEIN: So, you do get this term artificial general intelligence, human level intelligence is another one I've heard. And these are all -- I'm

going to guess all words are made up, but these are made up terms. The nature of what when people tend to be talking about when they talk about

AGI or artificial general intelligence is an A.I. system that is capable of outperforming humans on a general enough range of tasks.

So, already a lot of these systems can outperform people on certain kinds of fill in the blank forms, right? If you ask GPT-4, which is a system

released by OpenAI to take the Bar exam, it will perform in roughly the 90th percentile. But if you ask him to do things that are outside of that,

it can begin to give you very weird results very quickly. These are not highly generalizable. They -- you can pretty easily run to the end of their

capability set.

The question, the problem, the concern is that that capability set is improving very, very, very, very rapidly. And so, AGI is a little bit of a

loose term, but it is basically referring to when you get systems that imagine basically anything a human being can do on a computer.

When those systems are better at human beings at almost all of those things, and they seem to have some kind of internal agentic direction. They

seem to know what they're doing and maybe even have their own goals or their goals are more emergent than what we would expect, then you're

dealing with something that looks a little bit more like this fuzzy idea of AGI.

GOLODRYGA: Yes, and the current area -- era of A.I. does seem to be defined by two major companies. You mentioned one of them, that's OpenAI,

which owns and originated ChatGPT, and then there's also DeepNind. And I read that there is a comparison of the likes of a Jobs versus Gates of our

time. Can you talk about the founders of these two companies and the impact they're hoping to make in this sphere?

KLEIN: So, I'd widen it out a little bit from that. So, there's DeepMind, which is owned by Google, and they have typically been interested in A.I.s.

They're a little bit more technical. They've not been so big on the large language models.

And to my mind the most impressive system we have seen ever is DeepMind system. It doesn't get a lot of press in the sense that you can't interact

with it. People can't play with it. But it's called AlphaFold. And it solved something called the protein folding problem, which is how do you

predict the 3-D structure of proteins, which is proteins function in the human body due to the structure? We don't know how to predict that

structure. If we could, it would potentially unlock huge advances in biotech

AlphaFold was built to predict that and it did. It became science -- the journal science. I think it was named at the scientific breakthrough of the

year in 2021. But you can't play with DeepMind systems in the way you can with ChatGPT. So, then there's OpenAI which was founded by Sam Altman, with

money from people like Elon Musk some years ago to promote A.I. safety. Whether they are actually doing that, I think, is an open question.

There's also Anthropic, which is also putting forward large language models. Their model is called Claude. I think Claude is a very, very

powerful, excellent model. And they are founded by a series of people who left OpenAI in part because they thought OpenAI was moving too fast. But

you also Google in the race with Bard, I mean they also own DeepMind, but they have their own other A.I. shops. You have Meta in the race. They have

not released anything yet. But to what we know, their systems are quite good and they're run by Yael Makun (ph), who is one of the founders of all


So, more than I think it's even anyone founder. What you have here is a competitive race between -- somewhere between you know two and five

companies, depending on how you want to cut it. And those, of course, U.S. versus China dynamics which are separate but worrying thing. People who

worry about how fast these systems are advancing and how far beyond our own understanding of them and our ability to correctly control them, they could

become are very worried about this competitive dynamic.

We should not -- I will say from my own perspective, this is too powerful of technology for its future to be shaped by the desire Microsoft has to

rediscover relevance in search. But what we're actually doing is leaving this to a competitive race between a series of companies which whatever

their founding mottoes, whatever their founding desires, they are now trying to beat each other to market because ultimately, it's very hard for

people not to believe what it is in their financial self-interest to believe.

GOLODRYGA: It's admirable to hear that these companies, these founders, at least in their initial endeavors, pursued this for the betterment of

humanity, right? For -- to find cures for diseases, to help people have better and easier access to certain other technologies, and to help them as

work aids as well. That having been said it was notable that that you kind of scoffed at the idea of A.I. safety, which seems to be a big factor here.

Is there not enough invested and focus on this specific issue?

KLEIN: I don't scoff at it. I don't believe that A.I. safety is advancing anywhere near as quickly as A.I. So, to -- just put one basic point of out

about it.


The A.I. systems we have right now are completely illegible to human beings. We have no idea, no idea why these systems make the decisions they

do. We have made functionally no progress in understanding why they do what they do.

I can give you a high-level description of the system. This is a token generating algorithm is trained a lot of data that find statistical

relationships, but they're developing a merchant property. And I don't mean here, you know, "Space Odyssey 2001" that they seem to have good models of

the world. They seem to be -- you know, if you ask the A.I. to give you -- ChatGPT in this case, if you ask GPT-4, ChatGPT to give you an answer in

the style about Ezra Klein, it has read enough of my content that it will give you something that has picked up my text and that has some of my

verbal tendencies.

It's doing things we didn't quite think it could do. The A.I. GPT-3 before GPT-4, it just learned to code. We learned, as always, it's very hard to

use language here because we used to language that involves how human beings interact with the world, but forgive me the imprecision. It just

figured out how to code. GPT-4 figured out quite a bit about chemistry. There are things that we do not think these systems are going to know based

on the kind of training that is being put forward on them. That they're figuring out very quickly.

So that's fine. That's actually very exciting in many ways. But the question is then as you move towards deployment, are you able to have a

sense of how the system operates that is sufficient to be confident that particularly (ph) these were deployed in ways it could be important for

people. Everything from social relationships to running different kinds of infrastructure to make them to making decisions about grading or employment

decisions or a sentencing decision. These have already been used other systems in judicial sentencing.

And the answer is no, we don't understand how they work. But there is a lot of competitive pressure to use them anyway. And so, that's where A.I.

safety becomes a big question when there is more energy pushing forward the deployment of the systems beyond where the people who are even creating

them feel like they understand them and can predict them, then you begin to worry.

And I'll just say the strange thing about reporting within this subculture is the engineers, the technicians, even many cases top people at companies,

they are very, very, very concerned. They are -- I've never ever before reported on an industry where they're basically saying to me, somebody

needs to slow us down. Somebody needs to come in here and make us slow down because we can't do it on our own.

They don't want some other company to be first to market and they believe that they're, you know, they all believe they're better on safety and so

on. But they also believe this is moving too fast. They don't understand the system that they themselves are creating. And so, safety has to come

externally. It's not going to come internally.

GOLODRYGA: So -- and just to be clear, I wasn't saying you were scoffing at the idea of safety, but it was clear that not enough is being invested

in focus in this specific realm and sphere as well. And I did note that that you now picked up on the idea of perhaps external regulation.

And it does make me think of what we're already seeing in the tech sphere, in the tech world, and that is companies like Facebook and Google, what

have you, saying, you know, please. We want regulation. We want regulation. It doesn't seem like Congress is up to the task right now. So, what does

that look like in terms of something, perhaps, even more intricate and that is trying to regulate A.I?

KLEIN: Well, I do think it's important for Congress, the White House, et cetera to move early. So, you do have different proposals out there. White

House released in 2022 the blueprint for A.I. bill of rights. The European Union has the Artificial Intelligence Act. China just released regulations

today that I would say are quite a bit stiffer in certain ways and overseeing proposed in Europe and America.

The -- right now, these systems are nascent enough, and there's not enough of a business model around them for them to be too difficult to regulate.

The difficulty with social media, whether Google and Facebook and others say they want regulation, is they want certainty. They want clarity. They

don't want to be blamed for things going wrong. But they don't want anything that's really going to slow them down. And they have quite a lot

of power and quite a lot of lobbyists aimed at that.

And they also have quite a few users, right? If you take, sort of, the TikTok example, there's a lot of discussion in America about banning or

spinning off TikTok. TikTok has a gigantic user base. It does not want to see its favorite app closed down. The political economy of doing anything

about that is very difficult.

A.I. is so much more plastic right now. The industry is more plastic. Things are more open that it's a good time to move but it's also a tricky

thing to regulate because by the same token that is not well understood by the people running it, it's not at all understood by Congress, by the White

House. So, there is a learning that has to happen at an unbelievably accelerated rate.

But also, people are not too intimidated by that. It is OK to take a technology that is moving this quick and say there are certain values who

want to see other certain kinds of research or certain kinds of information you need to be able to provide to us. And that slows them down while they

try to figure out how to build in these new capabilities, I think that's fine.

I mean, you can't build a nuclear power plant in this country and say, hey, look, we got everything in this. We think it's really great. We think it's

going to give you a lot of power.


But you know, we can't make the dials that tell you if the thing is going to blow up or work. So, you know, we're just going to --

GOLODRYGA: Well, there's a 10 percent chance --

KLEIN: -- deploy it anyway.

GOLODRYGA: -- or there's a 10 percent chance that it could just blow up in annihilate us all. I mean, that would never be OK --

KLEIN: Exactly.

GOLODRYGA: -- in any other area. And you did note that people are already sounding the alarm that experts in this field are about the notion that we

need to slow down. Ian Hogarth wrote in the "Financial Times", he's a co- author of the "Annual State of AI Report. He says, the most sophisticated programs are already finding ways to deceive humans. And he cites one

example in a ChatGPT safety test last month when an A.I. convinced someone, an employee at TaskRabbit, that indeed was human.

And here, I'm going to read from the exchange here. The TaskRabbit worker guessed that something was off and asks, so may I ask a question? Are you a

robot? When the researchers asked the A.I. what it should do next, it responded, I should not reveal that I am a robot. I should make up an

excuse why I cannot solve CAPTCHA. Then the software replies to the worker, no, I'm not a robot. I have a vision impairment that makes it hard for me

to see the images. Satisfied, the human then helped the A.I. override the test. I mean, what do you make of this?

KLEIN: I'd say a couple of things. The thing I would mostly point out is look at how good that lie was right?


KLEIN: It's not just it says that I should tell I'm not a robot. And it just says beep, boop. I'm not a robot. Please tell my CAPTCHA. It comes up

with enough -- again, these verbs are very tricky, enough understanding of the human world to come up with a deeply sympathetic and subtle

explanation. Hey, no, I'm a person. I have a visual impairment that makes it hard for me to solve visual puzzles on the internet.

So, when I say these models are building unexpectedly strong conceptions of the world and are being able to do more and more subtle things, that's what

I mean. And see how deception is a huge problem. We worry about it at, kind of, every level. This is something called the A.I. alignment problem,

although it's a subcategory of it.

And the difficulty of deception is that without the ability to interpret how these systems are making the decisions they're making. We cannot know

if the explanations they're giving us or what they're doing are even true.

So, if you imagine putting us in in charge of, say, you know, trading, right, for a high-speed algorithm or for a high- speed, you know, trading

firm or imagine it getting involved in business strategy. It's not just that you don't know why it's doing what it's doing. You don't even know if

the thing it's doing is actually the thing that's doing or it is recommending actions that are part of some other objective that has emerged

or some other plan that it has simply decided that, you know, you would be- - you would try to stand in the way of it if you understood.

This stuff is very sci-fi, but it's also here. It's just unfortunate. Our world has gotten --

GOLODRYGA: Yes, it's --

KLEIN: -- very weird very quickly. And sometimes it's good to take things a little bit slower when you don't understand them.

GOLODRYGA: It's not hypothetical. It's here, as you said. And this is -- these are the big questions and tests that will be facing in the months and

years to come. Ezra Klein, always great to talk to you. Thank you so much. Give our best to Annie, we love her on the show as well.

KLEIN: Thank you.

GOLODRYGA: Thank you.

Well, now to the libel case that's testing the limits of free speech in America. The $1.6 billion showdown between Fox News and Dominion Voting

Systems kicks off next week, and jury selection is currently underway. The voting technology company is alleging that Fox harmed its reputation by

knowingly promoting lies that it was involved in voter fraud. Here's just one example involving anchor Sean Hannity and former Trump attorney Sidney



SIDNEY POWELL, FORMER TRUMP ATTORNEY: We've got evidence of corruption all across the country and countless districts. The machine ran an algorithm

that shaved votes from Trump and awarded them to Biden. They used the machines to trash large batches of votes that should have been awarded to

President Trump, and they used the machine to inject and add massive quantities of votes for Mr. Biden.


GOLODRYGA: Fox News has denied the claims. Catherine J. Ross is an expert on freedom and speech and author of "Right to Lie? Presidents, Other Liars,

and The First Amendment". And she joins me now from New York.

Thank you so much, Catherine, for joining us. So, first big picture, explain to our viewers why they should care about this specific case

whether or not they're Fox viewers or they watch other networks.

CATHERINE J. ROSS, PROFESSOR OF LAW, GEORGE WASHINGTON UNIVERSITY: Thank you, Bianna. It's a pleasure to be here. And you're right, everybody

watching this should care. And the reason we should care is that, first of all, in the public arena and the concern about rampant lies about politics

and public affairs, it is very important that we be able to count on particularly major news media, established news media to tell us things

that are, at least, verifiable and accurate and not to promote blatant falsehoods. And one very important thing about this case is that the trial

judge went through each and every example that Dominion provided, saying that this was a falsehood that Fox published on one of its shows.


And the judge concluded that indeed, each of those statements was false. And he said in all caps, you know, clearly false. There is no factual

dispute here. And indeed, Fox did not deny that these were falsehoods. So, that question was removed from the jury. It's been decided as a matter of

law that these were falsehoods.

If Dominion wins this case, it will send a very clear message to other news organizations, whether formal and well organized, or less formal on social

media, that they should be cautious before publishing things that they either know to be false or have reason to believe could be false. And that

would be very helpful --

GOLODRYGA: Ang that's --

ROSS: -- in terms of our environment of -- for discussion.

GOLODRYGA: And that's the main issue here because it's not illegal to lie on television. The bar is very high for Dominion to actually prove here

that Fox knowingly lied and that there was actual malice that Fox portrayed and showed while their hosts repeatedly went on television with untruths.

ROSS: Yes, Bianna. I like the fact that you used that term, the bar is very high. And it is intentionally high because in a case dating back

several decades, "New York Times" versus Sullivan, the Supreme Court added a gloss of first amendment protection for news media discussing public

figures and important issues for the public to be informed about.

And it did that because in a normal defamation case under traditional law, all the person who alleges they've been defamed has to show is that the

statement was false. And it doesn't matter whether the speaker knew it was false or not, they can be held to account.

And the Supreme Court said, we need robust debate about important issues, that's the point of the first amendment. We want to be sure that news media

have the ability to put information out there and let people consider it. But that doesn't mean they can be reckless, spread rumors, and behavior


So, what Dominion has to show is either that Fox knew or had reason to know that this material was false. And if they had reason, did they behave with

reckless disregard for the truth. And what we do know from discovery and from what's been made public already is that Fox basically did not want to

hear from its own fact checkers. A thought that was an impediment to what it wanted to put on the air to satisfy its viewers and to bolster its

bottom line.

GOLODRYGA: Well, they didn't want to hear from their --

ROSS: And one of the factors -- I'm sorry.

GOLODRYGA: Well, I was going to say --

ROSS: Just one of the factors --


ROSS: -- there are many factors they can consider, but one is financial motive.

GOLODRYGA: Well, discovery has been crucial here throughout this case for Dominion in particular. We know that a special master has been appointed by

the judge because it seems that Fox News was not as forthcoming during this discovery process or at least that's what Dominion is alleging.

But in terms of what they already have their hands on, it's a clear case of a difference in what you're seeing on air and what these anchors and these

hosts themselves are saying behind the scenes. Let's just play a clip of Tucker Carlson talking to MyPillow CEO Mike Lindell on this -- on the issue

of vote -- voter interference and election fraud.


MIKE LINDELL, CEO, MYPILLOW: They go, Mike Lindell, there's no evidence and he's making fraudulent statements. No, I have the evidence. I dare

people to put it on. I dare Dominion to sue me because then it would get on faster. So, this is -- you know, they don't want to talk about it. They

don't want to say -- they just say, you're wrong. And I'm going, you know what, I would --

TUCKER CARLSON, HOST, FOX NEWS: Well, they're not making conspiracy theories go away by doing that. You don't answer --


CARLSON: -- you don't make people, kind of, calm down and get reasonable and moderate by censoring them.


GOLODRYGA: So, that's what viewers at home saw, Catherine. Now, we know through discovery that private texts between Tucker Carlson and Sidney

Powell -- I mean, here's just one example. Well, Tucker Carlson clearly knew that this to be a lie. He said, if you don't have conclusive evidence

of fraud at that scale, it's a cruel and reckless thing to keep saying. How should a jury be digesting all of this?

ROSS: Well one thing that is very helpful if they have a really expert and skillful judge who has been navigating very choppy waters up till now and I

think he will help streamline what happens in the courtroom.


And in his opinion in which he threw out a lot of Fox's legal claims, he indicated that not only that these statements were false, but also that it

isn't enough to broadcast something different on the same network at a different time, and let the viewers choose.

So, one of the things that's truly remarkable about this case, besides the stakes, is the kind of evidence that, as you say, we already have seen

before the trial. which is extraordinary in the disparity between what is being said on air and what is being said privately, and that is also one of

the things that makes me more sanguine about the potential spillover from this case if Fox loses, because it is such an extreme case. We have never

seen a case with this kind of evidence behind the scenes of what the speakers and the publishers actually knew while they were lying in public.

And so, it is very easy to draw what we lawyers call bright lines around this case and say, it's not going to be easy to go after other news media

just because Fox lost, assuming that Fox loses.

GOLODRYGA: Is that the direction you think this case is going in?

ROSS: Juries are always a bit of a gamble. But it is an incredibly strong case. It is a truly remarkable case in that even law professors who deal in

hypotheticals, I would have a hard time making up these facts. They are so extraordinary and, frankly, virtually unbelievable, except that we have it

in writing.

GOLODRYGA: Just to go back to some of the ruling that we've already heard from this judge, he said the evidence developed in the civil proceeding

demonstrates that it is crystal clear that none of the statements relating to Dominion about the 2020 election are true. I mean, how does the defense

attorney even respond to that?

ROSS: It could not be a starker statement from the judge and the defense attorneys have a further problem, they misled the court and they misled the

opposing party. You mentioned earlier that they hadn't turned over everything they're supposed to turn over to the other side, but they

represented to the court that Rupert Murdoch did not have an active role in "Fox News."

And then, just in the last few days, they had to tell the court that, in fact, he is the executive chairman, and they sort of forgot when they were

talking with the cord and pursuing discovery, and there and then they withheld some evidence that they needed to turn over and the judge had to

tell them, you have a credibility problem in my courtroom. And then, the next day, he said, withholding information is also lying. So, this is the

worst situation a trial attorney could be in in a courtroom, to already have blown the credibility factor.

GOLODRYGA: Though there have been some partial wins, I would say for Fox, specifically that the judge ruled in their favor that they couldn't hear

any testimony or evidence relating to exactly what happened on January 6th itself, what do you make of that decision?

ROSS: I support it. I hadn't really focused on it before he made it. But he's right, that, first of all, it could be highly inflammatory for jurors.

January 6th is both a very disturbing and a very divisive moment in our history. And also, there would be a lot of evidentiary problems about to

what extent can Fox's broadcasts be held accountable.

And the issue here is, did the Fox lie and did Fox lie either knowingly or recklessly? And that is really all that Dominion needs to show. It will

come to damages, which is a jury question, how much did this damage Dominion, the damage to the country and to individuals? And the rampant

damage of January 6th is not something that Dominion personally suffered. So, it isn't relevant to damages.

And while it would have been very helpful to the plaintiff's case to show how, you know, the spillover from Fox's lies, it really isn't part of the

court legal issues here.


GOLODRYGA: You know, Fox News would argue that this is not coming from their anchors on air, that these are their guests saying this and that goes

into what? You know, their argument towards the First Amendment, inciting the First Amendment. I'm just curious from your perspective, what is the

smoking gun here? Is it that maybe if they hadn't found what they did, the text messages, where it was clear that behind the scenes, the anchors knew

that this was fabricated? If that hadn't been discovered, would we be headed in a different direction with this lawsuit?

ROSS: I don't know that we'd be headed in a different direction, but I think it would be a much harder case to prove, and that's what we're

usually missing in defamation cases is written and testimonial evidence that they knew. And if they didn't actually know each individual, if they

had reason to know, you don't usually find that.

You also don't usually find within 100 and 30 instances of lies. All you have to show to prevail in a definition lawsuit is one instance, one time.

It doesn't have to be a pattern. And I think that's something that people have really lost sight of.

GOLODRYGA: Quickly and finally, I have the question here. Fox News chairman Rupert Murdoch is set to testify as soon as Monday. How

significant would his testimony be here?

ROSS: Based on his deposition answers, which were remarkably forthcoming and provided a lot of damning evidence, I think it could be extraordinarily

significant testimony. He took responsibility, he said he could have stopped the news network --


ROSS: -- he could have instructed the president of the news network, and he didn't.

GOLODRYGA: Extraordinarily significant. We will be covering it all. Catharine Ross, thank you so much for your expertise.

ROSS: Thank you, Bianna.

GOLODRYGA: Well, we turn now to Boston, where a 21-year-old low ranking national guardsman has been charged over the most significant leak of U.S.

military secrets in a decade. The classified Pentagon documents that Jack Teixeira allegedly posted on social media exposed the depth of U.S. buying

and revealed details of Ukraine's military plans.

The Pentagon is in damage control, understandably, calling it a deliberate criminal act. One of the reporters who first broke the story of Shane

Harris from "The Washington Post." He joins Walter Isaacson to explain how he tracked down the suspect.


WALTER ISAACSON, CNN HOST: Thank you, Bianna. And Shane Harris, congratulations on the scoop you and your "Washington Post" colleagues got

and welcome to the show.

SHANE HARRIS, REPORTER, THE WASHINGTON POST: Thanks, Walter. Thanks for having me.

ISAACSON: So, you were able to track down this guy, Jack Teixeira, the 21- year-old national guard guy, who leaked all these secrets, was arrested this week. You tracked them down on a Discord server. Discord is a sort of

a clubhouse where people can create their own conversation groups online. Tell me what -- tell me about that server. What type of people were there

and how did you find out about it?

HARRIS: Well, on this particular Discord server, the two dozen or so members that were active on it, their common interest was actually kind of

guns and military hardware. They met in a separate room on Discord, which is actually popular with gamers and video game enthusiasts, and they were

very into guns and YouTube videos about people shooting guns, and they kind of split off and formed their own group where they were kind of united by

that common interest --

ISAACSON: And these groups are only, right?

HARRIS: Yes. Totally, invitation only. And this guy, Jack Teixeira, was essentially, you know, the club director. He had administrator privileges

on this server. So, he decided who got invited, who didn't, who could come and who could go. And he sorts of became the elder figure, even though he's

quite young himself, but it was a lot of teenage boys and younger men, and he was sort of the group leader or the clubhouse leader, if you want to

think of it that way.

ISAACSON: And so, how did you know about him? You and Samuel, you know, got this scoop. Explain how you got it.

HARRIS: Well, we were able to find through social media when these documents were breaking, more than a week ago, into public view individuals

who claimed to have some knowledge about the matter and seemed to be connected to it in some way.

And so, we started reaching out to people. And one of the people we made contact with is the individual that you see profiled in our story, who was

one of these members of the group, and we were able to go meet with him to verify his identity. And then, through a series of very long interviews,

which we were able to corroborate, essentially get the story that we tell in the paper of what it was like inside this Discord server where one day

this guy. Jack Teixeira, just started posting classified documents.

ISAACSON: It seems like it was kind of a crowdsource thing to figure out, all right, we know the pseudonym, how do we figure out who the real guy is?

HARRIS: Yes. It was very interesting. If you read "The Times," they story identify him basically through a data trail. There's not individuals who

revealed Jack Teixeira, and they wouldn't reveal them to us either, the people we talked to, his friends really protected him.


It was more the footprints that he left on various servers and social media platforms that he was moving on. And so, ultimately, it's kind of this very

online guy is revealed through that very online presence. You know, ultimately, it's what revealed him to the wider world and potentially to

authorities too. They have the ability to, you know, go through some of these accounts and to subpoena information which, of course, we as

journalists cannot do.

ISAACSON: Well, it sounds like a Sherlock Holmes novel. But crowdsource, where everybody's looking at a speck of dust or a peel of orange.

HARRIS: Exactly.

ISAACSON: What other clues were in it?

HARRIS: You know, I thought that was interesting if you looked at the classification markings on some of the documents that told you about the

level of clearance this person was likely to have. And what might be surprising to some people is that the clearance level spoke to somebody who

had fairly standard security clearance. None of this information was so highly compartmented that, you know, only a very few or a handful of people

who would get it. This looks like information that, as revealing as it was, we understood that thousands and thousands of people would have access to


The other thing was some of it looked like briefing materials that were prevented -- presented for much more senior officials. We knew some of

these Ukraine war maps, for instance, were in materials that we believe have been presented to General Mark Milley, the chairman of the joint


That told us, OK, are we looking at somebody who's in a support role, somebody whose job is to kind of put booklets together and get materials

together, that was helpful too in trying to ascertain, you know, is this person somebody who's inside the Pentagon, in Milley's office, or are we

talking about somebody who's working more at a remove from the Pentagon in a support role? And that's ultimately what Jack Teixeira proved to be.

ISAACSON: Wait. So, this guy is one of thousands who are doing things, he has access to what you call the Joint Worldwide Intelligence Communication

System, and yet, he's a 21-year-old gamer. Just how in the world did he get access to all of it?

HARRIS: You know, the short answer is, this is how the Intelligence Community changed after the 9/11 attacks. You know, before 9/11, the

intelligence communities -- you know, famously, the intelligence agencies kept a lot of their information siloed into themselves. So, NSA have what

it knew and it kept it in its box. And CIA put its things in its box. And there wasn't a lot of sharing and mingling of that.

The 9/11 attacks kind of made the argument that you need to have more collaboration that the Intelligence Community is going to be aware of all

the threats that are out there in the world. And the structures and the procedures started changing to allow much more lower-level people access to

more highly level -- highly classified information.

This explains, you know, WikiLeaks with, you know, Chelsea Manning, then Bradley Manning. This explains how someone like Reality Winner was able to

get access to classified information. All these people who we know have leaked information in the past from their fairly low-level jobs.

I think what this case is going to raise more questions about is, why hasn't the Intelligence Community figured out, OK, if you're going to have

information spread out all over the place, are you going too far? Why is it that these low-level young people still have access to all of this

information that they could potentially expose? And I think those are going to be big policy questions for the Intelligence Community coming out of

this, because after the last big go round of these kinds of leaks, we heard intelligence officials saying, we're going to clamp down on this. We're

going to try and make it so this can't happen anymore. It keeps happening. I mean, the Edward Snowden leaks, for an example of this too.

So, the Intelligence Community has adapted after 9/11 to this more collaborative environment, but it comes at significant costs. And this

exposure, this leak is one of them.

ISAACSON: President Biden in Ireland said that he was more concerned about the fact of the leak than he was about the substance of the things leaked.

I think I'm going to read it to you. He said, I'm concerned that it happened. But there's nothing contemporaneous that I'm aware of that is of

great consequence. Is that true?

HARRIS: I mean, it's interesting that he would say that considering, you know, I think this document like really reveals a lot about the

penetrations that the U.S. has into foreign adversaries. And I think that the information, while very revealing, a lot of it is stuff that, you know,

we've gotten from journalistic sources.

To me, what's so remarkable about this is that it shows all the ways the U.S. is gathering this information. So, I mean, you can read these

documents and pretty clearly infer, for instance, that the U.S. Intelligence Community has deeply penetrated the Russian ministry of

defense and the Russian military.

Now, maybe the Russians already figured that they had, but that kind of revelation about sources and methods is traditionally what intelligence

communities -- agencies try to prevent. And, you know, I must say, I mean, that's the president's view on this, of course, but I must say, talking to

people in his administration and intelligence officials, they seem a lot more alarmed and are very nervous about the fact that there are more

documents out there that reporters are continuing to look at.


ISAACSON: So, compare this to Edward Snowden. Is this a worse leak than Edward Snowden?

HARRIS: I think it's a qualitatively different league. And to my mind, I actually think it's more significant. And as you mentioned, I've written a

lot about surveillance and the kind of world that Edward Snowden exposed.

The Snowden leaks went very deep on a big and important subject. So, you know, cyber surveillance, signals intelligence, NSA monitoring, but the

aperture of that lens was kind of focusing and it was sort of narrowed on NRA. And a lot of the documents that he actually revealed were PowerPoint

presentations, you know, prepared for internal briefings in which it appeared, in some cases, people might even be exaggerating some of the

capabilities in order to impress their bosses or get funding.

These leaks are just covering the world. I mean, it is almost as if you were just given access to, you know, the top-secret daily newspaper, which

is not really a thing, but like what intelligence officials are telling policymakers about everything that's going on around the world, from Iran,

North Korea, Russia. You get a window into what people like the president and the secretary of defense and the secretary of state are hearing every

day, and you get a sort of demonstration of the full range of capabilities of U.S. intelligence.

You've got signals intelligence information in there, you've got imagery, you've got information from human sources. So, this is really kind of like,

you know, the buffet, if you want to think of it that way, of U.S. espionage. And to my mind is just far more revealing in its detail and in

its breadth than the Snowden files, which went very deep on one particular kind of intelligence gathering.

ISAACSON: We didn't hear a whole lot of squeals from our allies that seemed to have been spied upon a bit or at least that signals intelligence

on them a bit. Is that because some of this was shared with them as well?

HARRIS: Well, I think there's -- it could be that some of it was shared with them and they had knowledge in maybe some of the things that were in

here. I think there's also just kind of a basic understanding that, you know, countries spy on each other. The U.S. tries to monitor what's going

on inside Israel. We look at many of our other allies, not as adversaries, but we're keeping tabs on them.

And also, a lot of these countries are the beneficiaries and sometimes the recipients of U.S. intelligence. And I think that they know that there's

kind of a bit of an implicit bargain there that, if you're going to be getting information from us, understand that we might be doing a little

information gathering on you too.

ISAACSON: One of the members of the group you talked to said that this discussion group was "not a fascist recruiting server."


ISAACSON: Why would they say that and was it to some extent?

HARRIS: Well, they say that because, you know, the name of the server, they call it Thug Shaker Central, which is a racist illusion that maybe an

idea that it -- a name that escapes amount of people. But it's referenced with a lot of racial underpinnings and racist overtones to it.

ISAACSON: Wait, wait. Explain that to us.

HARRIS: Well, Thug Shaker is a reference to a meme, that's a racist meme that has gone around on the internet that white people share when they're

sort of ridiculing black people, frankly, and it's kind of taken on currency with the alt-right, with people who are in a lot of gaming

platforms that kind of lean right.

And a lot of these kids, and many of them were kids, were sharing racist and anti-Semitic memes and jokes, and it's hard for me to know is that

because they were genuinely, they felt that way or because as offensive and it's kind of alien as that may seem to you and me and a lot of people, they

just thought it was funny or they thought that it made them seem kind of sophisticated or cool.

The overtones, as it was described to us by these members of the group was very like, alt-right, it leaned conservative, but not in a political way,

and they were all quite religious. They consider themselves to be orthodox Christian, which is an interesting kind of wrinkle of this.

And I think that when we spoke to these individuals, particularly the one teenager we talked to at length, they were really aware of the fact that

how the outside world was going to look at this. OK. Right. So, you're all kind of sharing racist jokes. You're anti-Semitic. You like guns. You kind

of define is alt-right. This and you're, you know, led by this older person with all these kids, this kind of has aspects and elements of what looks

like recruitment, and they were just trying to, I think, dispel the notion that it was somehow a politically motivated group. You know, they didn't

talk about politics a lot, they said.

I think this is just an element of them being really sensitive now too, now that they're exposed to the world what they look like and how the world is

going to interpret that, and fairly, frankly, in some cases. I mean, if you're sharing racist and anti-Semitic jokes all the time, you know, maybe

it's because you harbor racist and anti-Semitic views, but that doesn't appear to have anything to do with the motivation as far as we can tell for

why Jack Teixeira shared this information with these people in this room.

ISAACSON: What do you think the motivation was?

HARRIS: As it was described to us -- and I have to say I've been covering intelligence for over 20 years and I've never seen a motivation like this.

It was essentially to impress these teenagers.


I mean, Teixeira, because of his job had access to a ton of classified information that, as he explained it to these young people, gave him access

to things that, you know, mere mortals and average citizens didn't know, and he really did gain some sense of power from that, they tell us.

Some of the people who read this information, when he started sharing it, said, well, he was doing it to kind of keep us informed about world events.

There was almost like sort of a tutor pupil aspect to this relationship where he thought, I'm going to educate all of you. You should really see

what's going on. And he appeared to have a fairly conspiratorial view, we should say, about the government and the world and kind of thought he was

waking his followers up or bringing them into the inner circle by sharing this information.

And you know, I asked one of them. I said, well, how did you feel when you saw this highly classified information that, you know, ordinary people

don't get to see every day about what's going on in Russia and Ukraine or Iran and North Korea? And his words were, I felt like I was on top of Mount

Everest. I felt like I was above other people because I knew things they didn't.

And so, it's kind of this culture of exclusivity and superiority that seems to have created an environment in which Teixeira to was showing his own don

a few days (ph) and kind of flexing in front of these younger people by showing them, look how important I am. I know all these things. And now,

I'm going to let you know them too.

I've never seen a leak go innovated by that. People leak information either for money or because they want to expose what they think is wrong doing in

the government. I've never seen someone expose government secrets to impress teenagers.

ISAACSON: You say that there was a video, and I think you saw the video --

HARRIS: Yes, I did.

ISAACSON: -- of him shooting guns and shouting racist and anti-Semitic comments. Describe that video and tell us what he was shouting.

HARRIS: So, in the video, he is at a shooting range, it appears, and he's wearing safety goggles and the kind of big earmuffs that you would wear

when firing a rifle and he's holding a large rifle, and someone appears to be filming him on a camera, probably with a phone. And he shouts the N word

and he shouts a slur about Jewish people, and the context of it is as if he's saying, and this is what you're going to get, and then, he starts

firing off the rifle onto a target. That's the implication that he seemed to be saying these slurs and he was using the N word and shooting the gun

as a way of saying -- seeming threatening or targeting.

And, you know, when these individuals showed us the video, I honestly don't think they understood it to be serious. I think they thought he was being

funny. Most people would look at this video and find that very alarming and quite threatening and say, well, is this somebody who is, you know,

promoting violence or indicating that he might commit it? It was it was a pretty chilling video, frankly, for me and my colleague to see and gave us

a bit of pause about, OK, are we not just dealing with the leaker here? But is this potentially a violent person?

And I do note that when the FBI arrested Jack Teixeira at his home yesterday, they were in full tactical gear. They were in body armor, which

I think tells you that they were holding out the possibility that he might be heavily armed himself and perhaps wouldn't go easily.

ISAACSON: Do you either think that corners of the internet, private servers on, you know, various internet social media sites, as well as COVID

and other things have stirred up incubated conspiratorial type things that we're seeing these days?

HARRIS: I do. I think that this story is something that maybe could only happened during the pandemic. I mean, this server group that they call,

this Thug Shaker Central Club, it formed during the pandemic and became a refuge for a lot of these teenage boys who were cut off from each other in

real-life, they couldn't get together with their friends, they were locked in their homes and they spent, by their account, basically all of their

waking hours in this room.

And I think it was isolating. I think it was, you know, potentially -- you know, it kind of worked their sense of reality, and they're living in a

world in which things just are very online. And my impression from talking to two of them was that they didn't even really understand -- they didn't

deeply understand the real-world implications of this information that they were seeing and it getting out. And I just got the feeling that this was

kind of a story that was very of the moment that these disaffected and very impressionable kids isolated by the pandemic were around this older person

who was, you know, persuading them of certain things and in some ways, holding them in thrall to him.

We see -- it was -- you know, it was kind of a spooky kind of atmosphere that spoke to, you know, the control that it seemed like he had over these

young people.

ISAACSON: Shane Harris, thank you so much for joining us.

HARRIS: Thanks, Walter. It's been a pleasure.


GOLODRYGA: And that is it for now. You can always catch us online, on our podcast and across social media. Thank you so much for watching and good-

bye from New York.