B (3:32)

And then so in my business, which is meant to be the truth and evidence and facts business, AI is a real danger because just as we had, you know, in the 90s, Jamie was saying, you know, it was the Internet superhighway, that was that quaint term. Then we have social media, which doubled down on all of that and started really, you know, hyping the fake news and distributing it at warp speed. Now, obviously, AI can do that at even greater speed and create these viral moments that we don't even know what's happening and we have no way to stop. I remember also, you know, for instance, I had a panel on my program when Sam Altman and OpenAI came out, and I had some really interesting experts on, some of whom were really doomsayers. AI is going to doom us as a species. It's going to take over, not to mention jobs and all the other things people are concerned about, and others who said, no, wait and see. AI has unbelievable possible applications in medicine and science and other progress. And as I said before we started, it's been very useful on a daily basis. When we do interviews as journalists, we shove it into a system, into an app, and we get our transcription in a minute or so, rather than in the hours it would take to physically transcribe, you know, a 25 minute or a half hour or an hour long interview. So in those issues, it's really useful. But as it becomes more and more sophisticated, and this is the thing that I can't get my head around, Jamie, this, this whole business of, of generative AI, which is apparently when this thing becomes more and more, if not entirely human, like in its cognitive abilities, and then we're screwed. I, I don't know how one has a, has a defense against that and what it actually means.

A (5:24)

The level that I'm worried about is what we talk about a lot, which is what is reality? You know, democracy can't work if you don't have an agreed reality. Democracy requires an agreed reality. So then we can debate what we want to do about that reality. If AI is used to undermine democracy by the authoritarian governments who want to undermine democracy, that's what Russia and China want to do. They do it every day. And in my last job in the State Department. I saw that happen and I saw it at a level where AI was just starting, where they spent billions of dollars in creating alternative realities to affect democracies in Europe, in the United States and around the world. Now here's where your point comes in. Just take the issue of translations. The Russians and the Chinese used to have a position or a line or an argument or a claim or a conspiracy theory they wanted to push. But it was requiring thousands of people to disseminate that information in all the languages. And remember the fam episode during the 2016 election where they hired these people in Macedonia and they Macedonians had, I don't know, 3, 400 people in a place to disseminate these conspiracy theories. They don't need that anymore. Now the dissemination can be done by an AI tool immediately. The translation can be done immediately. And here's where the hard part comes. And with generative AI, they can automate the warfare itself. They can teach a tool to generate information warfare using the data that big data companies have given the world, where everyone knows everyone's opinions or what they like or what they don't like. And then they micro target. And that's what I'm worried about, the combination of all these hyper capabilities put together in an automated machine where you don't need humans to create these information warfare targets tools.

B (7:32)

So you know, you've raised a couple of issues that also on the minds of those who are writing into us, how will AI impact facts and fact checking? Which goes to the heart of what we've just been discussing, because if you don't have empirical facts and evidence and truth as I like to call it, where do you start? And I think, oh, that's a big, big issue. Because even the social media titans who've been ordered to fact check, as we've seen in the last months, certainly since the Trump administration, Trump 2.0, have actually pretty much decimated their fact checking operations. So that's really, really dangerous. For all of those who get their news from social media, not just their entertainment, but who believe news exists on social media, that's really, really dangerous. And then Jamie to you, Stephen asks what diplomatic strategies are being developed or should be to, to address everything, including cross border AI threats like election interference? Stephen asked, and of course mindful that the State Department has essentially again denuded these particular tools and dismantled the tools that they had.

A (8:40)

Look, it's a painful reality for me to tell you that we worked really, really hard. Hundreds of people that I managed at the global Engagement center to try to get a grip on this problem. And we had two kinds of things we were doing in this particular area. One was to have active defenses to interfere with what Russia or China were doing to prevent them from succeeding in creating conspiracy theories in Africa or lying about truths and, you know, lying about realities in, in Asia. And we had these programs and they were working. We were able to intercept Russian operations before they got started and succeed in killing them. And we were able to undermine Russian RT parent company Rosia Savodna, which was using its, its tools all over the world in massive covert operations. And we put sanctions on them. Those capabilities in the State Department have been eliminated. We've talked about, you know, rfe, Radio Free Europe, Radio Free Asia, Voice of America, all that is what you might call offense. What do we say? But the defense has been eliminated. The offense has been eviscerated by cutting the budgets. And what I worry about is that there were ideas out there deal with things like deep fakes. Because remember, what AI can do is create this synthetic fake reality through sound, through video and through text. It used to be just text. Now it's all of those things put together to create fake narrative arcs that affect individual countries. And the machines are culturally sensitive. They know what works. So we were building an international system diplomatically where the Europeans, the Americans, the responsible countries, countries in Asia we're going to try to develop standards for making sure we knew when something was AI generated and when it wasn't, sort of a watermarking system. And so that everyone would know this is AI, this is not AI. And that would begin to address the problem. All those capabilities were wiped out. And the funniest, sad, ironic part was Marco Rubio, Secretary of State, suddenly began to worry about deep fakes. And the people that were working on it he had just fired. And so he couldn't get the help that he needed to prevent him from being deep faked, which he ended up being, and the capabilities to work on this because of this right wing conspiracy theory that all these people working on it were part of some censorship operation, which we weren't. We didn't ever work in the United States. It had nothing to do with our work. But they fired all those people. And now our capability is limited and we have to rely on large corporations like Elon Musk's corporation, like Facebook and Meta, who have eliminated their fact checking, who don't seem to care about this at all. And that's why it's so worrying. Without government and big data and big companies who have these powers coordinating their action. And at least in this case, in the AI case under President Biden, the big AI companies agreed they needed to have some regulation in your program. They admitted that you had them on. Some of them did. But we can't have that if there's a total breakdown in a recognition that this is a problem. And that's what I'm worried about.

B (12:11)

And under a lot of pressure and hard work, you know, a lot of great people around this here in the UK have brought in, you know, quite a lot of safeguards. And in Europe also they have more regulation. Certainly they have none in the United States. You know, you brought up a couple of things which is about warfare. Just so happens that I interviewed James Cameron, who is the great director of Terminator Avatar and all of those and who is making a film in the future and he's talking about it on the anniversary. He did anyway of, of, of Hiroshima. And we did our interview on the 80th anniversary of, of Hiroshima and that we're going to talk about and the danger of AI in, in warfare later in the program. So definitely stay tuned for that. We're going to take a break now and when we come back we're going to drill down on the general stuff. We just talk. It affects elections and remember 2024 was a year of unprecedented number of elections around the world and quite significant numbers of those and important ones were interfered with. And importantly, the US Vice President JD Vance almost gave his benediction. We'll talk about that when we came back. Actually, he gave his benediction when we came. When we come back.

C (13:34)

Ever wonder why you have insurance for your car or home but not your digital life? Meet Webroot Total Protection, your digital bodyguard that is built for real life. Webroot takes the guessing game out of cybersecurity so you can confidently browse, bank and be yourself online without the worry of hackers lurking around the corner. With Webroot Total Protection, you get antivirus that scans six times faster and takes up 33 times less space than the other guys. Identity protection with up to $1 million in fraud, expense reimbursement and 24. 7 US based customer support. VPN protection that hides your IP address, personal data and location from hackers and cloud backup with unlimited storage that works automatically in the background with plans for individuals and families. Webroot makes it easy to live a better digital Life. Go to webroot.com promo and get 50% off today. That's webroot.com promo to get 50% off today, live a better digital life with webroot, because peace of mind shouldn't be optional.

B (14:40)

So back we are with part two of our discussion in which we're going to focus on not just misinformation, but how AI turbocharges anything, including misinformation and deep fakes makes everything look so much more convincing. We thought social media went viral. This goes viral before social media has even had a chance to weigh in. So what has been the impact, certainly in recent European elections this past year? So let's take a question from Matthew on this topic. He says, how might AI influence upcoming elections, both in terms of voter behavior and public trust in the process, I'm actually going to add, and the effect of AI on the process and on the election result. Jamie, there's been some, some examples. I know you want to dive in.

A (15:27)

Romania is the great example of how this all comes together. So there was a clear Russian effort to interfere in Romania's presidential election. So what happened in Romania was they created fake bots. Instead of having people at farms, they had bot farms where tens of thousands of bots suddenly created a grass movement in favor of this candidate, Georgescu, and they brought him from zero to being the leading candidate in the first round of an election. The intelligence community of Romania figured out what the Russians were doing because they used TikTok and they used Bitcoin to pay people to actually manipulate. So it was a classic case of interference. And when the Romanians canceled that election because it was a poison chalice, as the Constitutional Court president declared, it became controversial. And because Georgescu was a right winger, because he was saying things that the Trump team, including Vice President Vance, liked, they challenged the cancellation of the election. Now, in the end, after a lot of incredible work, the people of Romania stood up and responded. But it caused a lot of controversy. Conspiracy theories will be discussed in Romania for decades. But in the end, the Romanian people refused to be manipulated by Russia. And it backfired because the correct information became available. And the Romanian people voted for someone who had agreed that the election needed to be canceled, but didn't agree with how it was done because explanations weren't sufficient, which is a fair point. But again, we can't have these elections interfered with without damage because the Russians had a plan. They wanted to turn Romania against NATO and they almost succeeded.

B (17:24)

It is remarkable because, you know, after that first presidential candidate and the process was exposed, then you said they did another runoff, another election, and there was another far right, populist, you know, candidate, and the Same kind of interference happened on his behalf. But as you say, he didn't win. And the centrist mayor of Bucharest did win fair and square. But as you mentioned, the U.S. i mean, and J.D. vance, I was in Munich. It was at the Munich Security Conference when he took the. Oh, God, it gives me chills just to remember it, frankly. He took that example and the example of the far right in Germany, too, AFD alternative for Deutschland. And he basically said by trying to expose the truth about them, he said Europe was being undemocratic.

D (18:13)

The threat that I worry the most about vis a vis Europe is not Russia, it's not China, it's not any other external actor. And what I worry about is the threat from within, the retreat of Europe from some of its most fundamental values, values shared with the United States of America. Now, I was struck that a former European commissioner went on television recently and sounded delighted that the Romanian government had just annulled an entire election. He warned that if things don't go to plan, the very same thing could happen in Germany, too. Now, these cavalier statements are shocking to American ears. For years we've been told that everything we fund and support is in the name of our shared democratic values. But when we see European courts canceling elections and senior officials threatening to cancel others, we ought to ask. Ask whether we're holding ourselves to an appropriately high standard.

B (19:14)

Imagine that he lectured the European establishment at the Munich Security Conference as being undemocratic by silencing the votes and the voices of candidates who'd been overtly and evidently pushed by Russia and this massive disinformation campaign. So that was a moment of, of real fear for everybody because they thought, oh, my God, if the Americans are going to side with this, well, we've got, you know, we've got no hope. And then, Jamie, why do they do it? So the. Why does Russia, China and Iran do it? I mean, basically, because they have their own agendas, as you said. So one of them is about NATO and probably the other is they think they want to. Well, they do want to get clients who are manipulative, manipulative and, and, and for them, for them. And they also want to really damage democracy. So that is another issue, right, this kind of interference and, you know, showing democracy or trying to show it in a really negative light. They're interested in actually chipping away at the value of democracy, not to mention in Ukraine, which is the point of this awful war by Russia.

A (20:22)

It's not an accident that it's authoritarian countries that are doing this. And even if their candidates don't win, they get to undermine democracy by causing tension, by causing dissension, by causing anger, by eliminating the societal trust that is needed for democracies to be as successful as they should be. So what happens here is they find every single issue. And this is where the AI part comes in. AI can help using big data to figure out what are all the points of tension in a society, the hundreds of issues that generate tension and dissension and anger among people. And then they use those tools against us by using AI to identify those areas, using it to disseminate conspiracy theories or messages in all the different languages instantly. And then they can automate it so that it's happening all the time. And until democracies figure out a way to create what's called societal resilience, making sure a society is resilient against this interference in our way of life. The Russians and the Chinese, even if they don't win the particular election, can damage the democratic trust that's needed for us to succeed. Remember, democracies are more powerful when people believe in their leaders, even if they didn't vote for them. They have to believe in them. They have to believe they're legitimate. But if this trust is broken, if they, those who lost the election think the other side cheated, or those are, the anger is created. Remember, after an election is supposed to be a moment when one side concedes and says, okay, we lost fair and square and we'll fight next time. That's what used to happen in our countries.

B (22:10)

Remember in 2000, you know, Al Gore.

A (22:13)

Did, did immediately, and he conceded, even though it was a pretty hard run thing and it was a real question mark. And many people were angry at him for conceding, but he did it because he believed in democracy and American democracy, and he did the honorable thing. Where are these honorable politicians now? Instead, we have people who just have their ego so wrapped up in these elections and their actions, they only care about themselves. And that selfishness is what the AI tools and the information warriors of Russia and China are playing on. And that's how AI is damaging our democracy. And that's why I'm on the dark side theory of this, that yes, it'll do all these wonderful things for medicine and translations and transcripts, but look what it's doing to that trust.

B (23:01)

Yeah. And you know, obviously we have to mention clearly that January 6th was the apogee, if you like, of that kind. And then it happened in, you know, in Brazil afterwards too. Bolsonaro tried to Question the legitimate election of Lula and there was a major standoff and Russia must be licking its chops all over that because wow, did it work? You know you were mentioning before the people, the hundreds of people in Macedonia. So in these articles it just points out something we have to internalize and we are that the so called troll farms which these people had been so successful and they, they kind of controlled it and it's described as, you know, they, they were more quality. Now the bots are more quantity and faster and they are, you know, gaining ground every single minute. And just for my case, you know, in India they created an artificial news anchor entirely AI generated news anchor. And that, that's a scary moment. So they can impersonate any of us as well. And that's very, very scary. And as you say there's, I guess our companies have not yet come up with a way to make sure this doesn't happen and protect against that. So these are all things to talk about and to think about. But shall we go into take a break again and when we come back I think we should pick up on what we discussed in the first part. And that is what happens when AI gets hold of the levers of war making and warfare.

C (24:35)

Ever wonder why you have insurance for your car or home but not your digital life? Meet Webroot Total Protection, your digital bodyguard that is built for real life. Webroot takes the guessing game out of cybersecurity so you can confidently browse, bank and be yourself online without the worry of hackers lurking around the corner. With Webroot Total Protection you get antivirus that scans six times faster and takes up 33 times less space than the other guys. Identity protection with up to $1 million in fraud, expense reimbursement and 247 US based customer support. VPN protection that hides your IP address, personal data and location from hackers. And cloud backup with unlimited storage that works automatically in the background with plans for individuals and families. Webroot makes it easy to live a better digital life. Go to webroot.com promo and get 50% off to that's webroot.com promo to get 50% off today. Live a better digital life with Webroot because peace of mind shouldn't be optional.

B (25:41)

All right everybody, we're back and we're going to talk about the most dangerous of human endeavors and that's warfare. The human factor has to be involved. Otherwise you just to have, have Star wars or you know, video games, war games in the, in the right. So I Was really fascinated after the 12 day war. It's not a war. A 12 day strike on Iran by Israel and the US and then the retaliation by Iran with missiles. There was a lot of AI going on there. There was a lot of fake so called ordinary civilian testimony, especially in Israel because they're, they're good at this stuff too. And they're very high tech sector Israel and there was a lot of that. Do you rem that the people claiming this victory and that victory and this is where that missile landed and that. And it was quite scary, actually.

A (26:33)

Terrible. Remember, we all, many of us use the technology ways to get around to drive. That's an Israeli technology. Okay. That was invented in Israel. They are really good at this stuff and they have technologies that have been proven to be, you know, essentially spy tools that have been given to authoritarian governments. So the Israelis have a lot of explaining to do about this. Their role in undermining. Remember Israel's great claim was that they're a democracy. Well, if they're creating tools that are killing democracy and denying reality, that's not something they should be proud of. But yes, in warfare, AI deep fakes can really do damage because they can create false narratives as you're describing, where people claim things happen that didn't happen. And remember the Iranians are pretty good at it too and they, they do some of this as well. But in the end I just have to believe since I know that even when the Chinese US relationship was as bad as it was under Biden and it was very bad for a while there, really in the first year and a half, it took a while to stabilize on this subject at least there were some conversations between responsible officials on both sides to sort of make clear that when it came to the ultimate weapon, the nuclear weapons, that nobody was going to be stupid enough to transfer decision making power, launch power to a machine the way it happened in the famous Terminator movies by James Cameron that you had interviewed. So I think at that level we're going to be okay. But where you're talking about is where AI interferes with these conventional wars which now remember are drone generated wars where drones are doing the killing and machines are aiming those drones, machines are targeting those drones, machines are assessing the damage from those drones and if you get that all wrong, you can really make some terrible mistakes in conventional war.

B (28:34)

Exactly. But look, let's focus a little bit because as you said, you got your start in public service because of the terror and the fear, particularly under the Reagan administration, of some kind of nuclear War. And many of us as students, we read all the, you know, horror stories around it. We marched against nuclear war. We, you know, really were pretty scared about the whole thing. It hasn't happened yet, thank God, but there's been some near misses. I was again talking to James Cameron, who's about to do another big, big issue, and it's about nuclear war and it's going to be about Hiroshima. So what he said to me, and I think it's interesting, is that the narrative counts. People have forgotten what actually it looks like. Remember, Jamie, in the 80s, there was a TV film called, was it called the Day After? Something like that.

A (29:23)

Exactly the day after, yeah. And you remember, Christiane, it was so powerful, its impact on ABC News, on the country, that Ronald Reagan, the President, insisted that ABC News give television time to Secretary of State George Schultz to come on TV after the program, program to tell the American people, no, no, no, this didn't happen. This isn't real. This isn't going to happen. Calm down. That's how palpable the fear was back.

B (29:52)

Then, which is incredible. But that shows what we're lacking right now, and that is storytelling, whether it's filmmaking or reportage or whatever it is that focuses people's attention on what's ahead. So he said, yeah, I did. The Terminator did that. And, and I'm glad that we put it out there because it did have an effect. As you. I'd forgotten about George Shultz. I don't even know whether I was in the US at the time. Maybe I was, but I think you.

A (30:21)

Were based on the timing, because I remember it was like 1983.

B (30:24)

Like 83 or something.

A (30:25)

Yeah, 82. 83.

B (30:27)

Yeah, yeah, yeah. And I guess I'd just gone from college to CNN at the. Yeah, in 83. So. So that's really important. And then also in this interview was Ernest Moniz, who you also know, Jamie, the former former Secretary of Energy for the US Himself a nuclear physicist. And he was also instrumental in the very technical parts of the negotiations between the US and Iran over the nuclear deal, the JCPOA, back in 2015, which then Trump pulled out of. And I asked him, I said, what about AI and the military and particularly of nuclear warfare? And he said, well, look, as you just said, the presumption and the declaration has been that there will always be the human factor. There will always be a human in the chain of command, in the link of, of that chain. But you don't know what, what if that doesn't actually be a Fail safe, a fail proof system. What if mistakes are made? And he said in the era of AI, the decision making process is so contracted. The decisions that used to make maybe, I don't know, a day or let's say hours, and even then it was minutes. Yeah, minutes and seconds and you could have a terrible accident. Oof.

A (31:42)

Let me give you a couple examples that are very chilling and reflect the greater humanity. There was a Russian who died, I think about a year ago who was crucial in helping Russia avoid responding to a phony fake mistake in the system where they thought a nuclear attack was coming and they had to decide whether to launch a retaliatory strike. This Russian insisted it was a mistake. And in Russia that's not easy to do because you can really suffer terribly if you challenge the system. And he was a hero, a real hero that saved, you know, arguably humanity. There was a similar case in the United States where there was a flock of geese that was mistaken for a set of missiles. And again, individuals didn't respond to indicators that these attacks were coming. And you're absolutely right, with AI, those indicators can be that much more realistic, real, that much more persuasive using some spoofed device. Look, we've all watched James Bond movies where some terrible third party actor, rich man billionaire in Austria gets all the technology and the tools to try to convince Russia and the United States to attack each other. And with AI, and with this penetration and this dissemination and this narrative arc that can be created by machines, machines, you could imagine the real pressure being placed on presidents in both countries to believe that something real was going on when it wasn't. Because they are increasingly relying on machines and increasingly doubting, you know, in an authoritarian system, remember the loyalty to the guy above you is what prevents you from getting killed or sent to the prison, even if you know something is wrong. And the more authoritarian it is, is the more dangerous that refusal to have the human factor. So we have to have the human factor. And if we don't someday, I really do worry that that could happen. But for now that's not my biggest worry. My biggest worry just by all Mike, is democracy being slowly in a daily basis through attrition, damaged the trust in all of our democracies so that we can't function successfully. And these authoritarian, like China, you know, they want to create an Orwellian world where they can watch what everybody does, decide who's good and who's bad and penalize them. And they are long term thinkers and they have the power of totalitarian control that is not undermined by democratic weaknesses that that AI can, can, you know, exploit.

B (34:29)

So while I just remind you that you're going to have the Rich Man Billionaire Austria coalition coming off to you. We didn't mean anything. He didn't mean anything.

A (34:39)

That was just from a movie. It was just from a movie. A Bond movie, I think it was. Oh, no, actually, it was the. It was the movie.

B (34:46)

I can't believe you're doubling down.

A (34:48)

Anyway, it was a movie. You didn't mean it was a movie. I didn't mean it.

B (34:53)

So thank you everybody for listening. And don't forget to follow the feed so you never miss an episode. You can always listen to Christiana Monster presents the X Files with Jamie Rubin on Global Player or wherever you get your podcast. And of course, you can always view it when it drops on YouTube. Okay, bye X. Goodbye.

A (35:10)

Bye X. Bye.

B (35:15)

Especially in the nuclear age. Whoops. Phone's going. So we're going to take a break.

A (35:23)

This is a Global Player original podcast.

C (35:27)

Ever wonder why you have insurance for your car or home, but not your digital life? Meet Webroot Total Protection, your digital bodyguard that is built for real life. Webroot takes the guessing game out of cybersecurity so you can confidently browse, bank and be yourself online without the worry of hackers lurking around the corner. With Webroot Total Protection, you get antivirus that scans six times faster and takes up 33 times less space than than the other guys, identity protection with up to $1 million in fraud, expense reimbursement and 24. 7 US based customer support VPN protection that hides your IP address, personal data and location from hackers and cloud backup with unlimited storage that works automatically in the background. With plans for individuals and families, Webroot makes it easy to live a better digital Life. Go to webroot.com promo and get 50% off your today. That's webroot.com promo to get 50% off today. Live a better digital life with Webroot because peace of mind shouldn't be optional.