Ed Vitagliano: And welcome back. Ed Vitagliano sitting in for Tim Wildmon today, joined in studio by Wesley Wildmon and Tony. And now Steve joins us, living up to his name. Very nice paisley shirt again.
Steve: When you go paisley, you go full paisley.
Ed Vitagliano: I go full paisley.
Steve: You have some nice shirts.
Ed Vitagliano: See, you have a life. Represent. When the season rolls around, do you have any Christmas paisley shirts or anything?
Steve: I don't know. Christmas colors, maybe.
Ed Vitagliano: That's something to do this afternoon.
Steve: Yeah, I'm going to be shopping for shirts.
Ed Vitagliano: Heather, I'm spending some money, but it'll be well worth it. You wait until you can check me out in my new Christmas paisley shirt.
Steve: All right.
Ed Vitagliano: All right, take it off.
Steve: So I know you guys were talking a little bit about the outages and the meeting up on the Hill. Did you guys tease the clip that I was going to bring in?
Ed Vitagliano: We forgot.
Wesley Wildmon: I don't think we did.
Ed Vitagliano: We did forget.
Steve: I hope I'm coming in.
Ed Vitagliano: Why don't you tease it real quick?
Steve: I think they're just coming because of you, Steve.
Ed Vitagliano: You know what? I need them for probably fifteen minutes.
Steve: Well, I want it to look.
Ed Vitagliano: Yeah, go ahead.
Steve: Okay, so we've been talking about all these outages, the FAA ground stop, Microsoft 365, and all these things. They are having a meeting on the Hill right now with Secretary of Homeland Security Markwayne Mullin, the deputy director of the FBI, and the national security adviser. A Republican from Mississippi, Michael Guest, asked specifically about this and whether they're related, and I want you to hear the replies.
You are going to hear first from Markwayne Mullin, then Jay Clayton, who's the director of national intelligence, and then Christopher Wray, who's the deputy director of the FBI. They are all addressing whether or not all these outages are related to cyberattacks.
Michael Guest: Are the outages that we've seen within the last week, do you have information that they are related? And then also, a second follow-up question: Do you believe that they pose any threat to national security?
Markwayne Mullin: We have seen nothing that would indicate that they are related. As the secretary said, after doing a great deal of due diligence inside the government, with the private sector, and with people in the private sector whose job is cybersecurity, we have no indication that these incidents are related or that they're coordinated. We also have no indication that they're the result of a state-sponsored malign actor.
Jay Clayton: We have our cyber action teams. Our cyber experts have been engaged from day one on all of these incidents very heavily. And I want to be very clear on this: At this point, there is no indication whatsoever that there is any cyber malicious activity related to any of those incidents.
Steve: All right, so today Apple is reporting that an internal software update caused disruption to Apple Music, the App Store, AppleCare, and iCloud. Marriott International users are having problems signing into their website. So we're being told that's all coincidental. Do you believe that?
Ed Vitagliano: Well, I talked to Bob again, who is kind of our expert on AI, and he says this has all the trappings of what happened a couple of months ago when Anthropic's AI bot got out of its sandbox, which is basically a place where it's supposed to be isolated from the outside world. It escaped and did a whole bunch of malicious damage. He says something similar seems to him to be going on here.
Steve: Are you saying that this could be what we're seeing, kind of an Anthropic-out-of-the-sandbox-type attack?
Ed Vitagliano: Yeah, I think it's all suggesting that. Now, I don't know that that's true, and I don't know that Anthropic, OpenAI, Google, or anyone else, including DeepSeek, is involved in any of this. But it seems as if, months after the initial acknowledgment that things got out of control, we began to see that it was really out of control, far worse than they initially established.
With a flurry of activity like we're seeing today, this suggests that, yes, indeed, things are a bit out of control, and that we need to be very concerned and very active in doing whatever we can to get it back in the sandbox.
Ed Vitagliano: All right, Tony, I want you to explain a thing or two here, because when I hear what Steve was saying and what Bob McGinnis is saying, Steve does a lot of stories on this, and you get to the place where you use the lingo of companies and platforms that are using AI, but most of us don't use that kind of language.
When both Steve and Bob McGinnis were talking about Anthropic escaping out of the sandbox, the AI agent, I think of Planet of the Apes, the first of the trilogy from a few years ago. These intelligent apes got out. They escaped from captivity. That's what it sounds like. What does it mean to say an Anthropic AI agent escaped from the sandbox? Bob McGinnis is saying some of what has happened here recently may be another example of that. Please explain it as if our listeners and I are listening.
Tony: Or as if Wesley is listening.
Ed Vitagliano: Or if Wesley is listening, explain it so that we understand what Bob is talking about.
Tony: I don't know. I'll try to come up with a metaphor.
Ed Vitagliano: You already gave me the metaphor.
Tony: Okay, sandbox is a metaphor.
Ed Vitagliano: Sandbox is a metaphor. Kids play.
Tony: Anthropic, in order to continue training their AI and improving it, obviously does tests with it. You put it in what you call a sandbox, where you give the AI freedom to do tasks that you normally would not want it to do in the wild or out on the internet.
One of the trainings they were doing with this AI was trying to see how proficient it was at hacking other systems independently, using just prompts and commands. What happened is that Anthropic claimed the AI they were training got out of the sandbox, escaped the gorilla, escaped from the cage, and started tearing things up in the lab.
Ed Vitagliano: Okay, right, and so this AI got out and was on the internet.
Tony: It was on the internet. It was doing real-world problems, causing some issues. They said it didn't do anything detrimental, but I think what is interesting about what Bob McGinnis is saying is this: I'm not saying this is what it is, but these attacks, or not attacks, these crazy coincidences could be caused by AI.
I'm not so sure about wild AIs that are just doing their own thing, but AI could be used by actors. You could just set an AI loose and give it commands and instructions, and it would continue seeking out these systems, seeking out vulnerabilities, poking and prodding. If it finds something, it's going to go in and exploit that weakness.
Some of these outages that we're seeing may indeed be caused by some of these companies rushing to patch vulnerabilities. They are rushing so fast that they are breaking things. But you're just asking a lot of me to believe the government in this instance, that these aren't related at all.
If you have a line of cars, and we all have the same car, and Brent's transmission goes out, that's weird. That stinks for Brent, but we're sorry about Brent. Then Steve calls me the next day and says, “Hey, my transmission went out too.” Maybe we need to look into that. Then I call the next day and say, “My transmission just went out.” At what point do you say, “Maybe there's an issue with the manufacturer”?
You would take that same logical course of action and apply it in this example. But it's like the government, maybe they don't want to cause panic, refuses to say that these issues are connected. Those are critical systems: the FAA, major banks, the NSA, Microsoft. Microsoft was out yesterday, right?
Ed Vitagliano: You're just asking a lot of me to believe you when you say they're not related. Wesley, what do you think? Am I crazy?
Wesley Wildmon: I had two questions for you. Is it possible? No, no, I know it's possible. Do you think that Anthropic recognized a hole that was left intentionally and then escaped out the “don't go out that door” sandbox?
Ed Vitagliano: Back in a couple of schools, you're asking right now. When did this happen?
Wesley Wildmon: The Anthropic thing happened a few months ago, probably back in the summer. But here's my question: Could it have gotten out on purpose, or could it have been allowed out? Were there any consequences?
Tony: I think with the Anthropic situation, you have to sift through what companies will do sometimes. Anthropic is not the only AI company that has said, “Oh, no, our AI is so advanced that it is able to get out and do these things,” but it never actually does anything really, really damaging.
Ed Vitagliano: They would be held liable for it.
Wesley Wildmon: That's what I was getting at. Would it not be beneficial for the company if it does get out, and then they can learn from it?
Tony: Well, I don't think that's one of the things you said.
Wesley Wildmon: Well, there's a bright side. We are at least learning what it can do and how to defend against it.
Tony: But they said it was an accident.
Ed Vitagliano: I will say this: I opened by saying Apple just reported that it disrupted services, and Marriott, on any other given day, unrelated, if Apple services were disrupted and people were having trouble with the Marriott website, no one would notice except for Apple and Marriott customers. It certainly wouldn't come up on this show.
Steve: Right.
Ed Vitagliano: But because of all the things that have happened, now we're just ticking off different things.
Tony: Yeah, I was just trying to tickle the ears of our conspiracy listeners here.
Ed Vitagliano: Well, they're ready.
Steve: Some of them may be in this very room.
Tony: Some of them are in this room.
Ed Vitagliano: One of them.
Tony: I am looking in the mirror to admit that.
Ed Vitagliano: That's why I use the example of the car. So, Steve, if Brent's transmission goes out this year, and a couple of years from now Steve's goes out, and mine a few more years after that goes out, we're not going to think, “Maybe the manufacturer had a problem.” But if they go out the same week, one day after the other, I don't know, it looks kind of coordinated.
And part of the reason, by the way, is that Tony's degree is in computers, and he heads up our IT department. So we're asking him questions for a reason.
Tony: Did he pay for it?
Ed Vitagliano: Did he call us to agree?
Tony: Well, in part, I did. I ate his food and my leftovers when he was still living at home.
Ed Vitagliano: So that's why we're asking Tony questions.
Tony: As these AI platforms and companies are trying to figure out AI, they put it in a sandbox. These AI agents learn, right? These are learning programs, and that's why they're testing their abilities inside that sandbox to see what they're capable of.
The question raised by Bob McGinnis is this: If it's not bad actors and it's not just a coincidence, then it may be possible for these AI agents to have become so adept at learning that they may be doing these kinds of things.
The testing that was going on inside the sandbox was to see if they could figure out how to bring down major U.S. institutions like banking, or somehow shut off the power grid, which is why I think Steve mentioned gain-of-function research. That's what the expectation is with how we got worldwide COVID: They were experimenting to see what happens if you change the DNA of something.
Ed Vitagliano: So is it possible that that's what was happening?
Tony: Yes, and this is what makes me lean toward a coordinated, intentional effort. We don't know what's happening in China. Maybe China's banks are going through the same thing. It's hard to know what's going on with the CCP. Maybe there's an AI agent. Maybe they have a similar allergy.
Ed Vitagliano: With bats?
Tony: With an AI.
Ed Vitagliano: With bats.
Tony: There's no way to know what's going on in Russia or China. Maybe they're experiencing the same issue from rogue AIs. If we got reports of that, I would say, “Okay, maybe rogue AIs are just out there doing this to anybody and everybody.” But we're only hearing about it in the United States and in European countries. And who is opposed to the United States and European countries?
That's what makes me feel like it's coordinated and intentional.
What makes AI so dangerous is this. I'll use another metaphor. I'm giving lots of metaphors to help you understand.
A normal hacker, if I'm going to use this building as an example, is going to try to get into the building. He's going to try the door, then he might go to the other door and check that door, then check windows one at a time. That's what normal hacking looks like.
But with AI coming onto the scene, now you multiply and duplicate the hacker. You multiply him by a thousand and send them all to the building, saying, “Get me in the building.” What are they doing? A thousand people are checking windows and doors. They start digging under the building, doing whatever they can to get in. They might climb on one another's shoulders and go into the roof.
That's what these AIs are capable of. They don't stop. They don't stop learning. They don't get tired. They don't go to sleep.
Ed Vitagliano: There's your answer, or something like that.
Steve: I appreciate that.
Ed Vitagliano: I thought AI existed just to give me a dancing cat playing a fiddle.
Tony: Like an image?
Ed Vitagliano: Yeah, it's doing something. It could be doing something far worse.
Steve: All right. So in this last election, we put a whole new raft of Democratic socialists in office, and they are driving the agenda. They are upping the ante when it comes to antisemitism.
For example, Zohran Mamdani was elected from New York, and Representative Melanie Stansbury was elected from, I believe, Colorado. They are demanding two things. They want the right of return for Palestinians to Israel, or what they would call Palestine, and they are looking for reparations from Israel for land they say was taken in 1948.
Ed Vitagliano: Now, this is a new wrinkle.
Steve: It is. It's upping the antisemitism angle.
The right of return means that whatever ancestral land your parents or grandparents were kicked out of, whether in Jerusalem or Tel Aviv, you have the right to come back and claim that property, kicking the Jews out. You would have a whole lot of Arabs and Jews without fences. It would really mean the end of the Jewish state, because it's kind of hard to coexist like that. They don't get along well without fences.
Then they would have the same problem with reparations that we have here with reparations: Who gets them and who doesn't? But this is a new wrinkle, a new antisemitism that the Democratic Socialists of America are prodding the Democratic Party to take.
Ed Vitagliano: All right, Steve, let me ask you this. I've had some text exchanges with friends and friends of this ministry. They don't like the label of calling everyone an antisemite who might want Israel to explain some things about the war in Gaza or have those kinds of questions.
They say that doesn't make me antisemitic. There are people in conservative circles who believe that Israel dragged us into war with Iran, promising that if we did, there would be regime change. If that is true, they would be critical of the Israeli government without necessarily hating Jews. We could go on and on with these examples.
The Democratic Socialists of America, though, clearly seem to be antisemitic. When you talk about the right of return, in other words, Jews having to give up their property and move out, and paying Arabs for land that was rightfully given back to the Jews in 1948, that's not just questioning the government. That is “from the river to the sea” type stuff. It would result in the murder of many Jewish citizens, which we saw from Hamas in 2023. You're basically placing a whole bunch of terrorists right in the middle of them.
Steve: I think it's fair to say you can question Netanyahu, his role, and the Israeli government without that being antisemitic.
Ed Vitagliano: Do you think that's antisemitic? I'm polling the room.
Steve: No, I don't think it's antisemitic to say, “Can you explain whether you were behind the bombing of Iran? Are you pressuring us to go back again and do it again?” Those are policy issues that are open for discussion, and I don't think that is antisemitic.
Wesley Wildmon: I like the way you set up the question. I would fall into the category of people you've talked to. I also like his answer. I don't have anything to add to this.
Tony: The litmus test for me has always been this: If I would ask the same question about U.S. geopolitical policy, then I should be able to ask the same question about any other country's geopolitical policy. If I had questions or concerns about something the United States was doing globally, the litmus test is that I can ask that about any country, including Israel, Mexico, Canada, or whatever. The list is open to everybody.
But what the Democratic Socialists of America are advocating for, as Steve put it well, is more than just asking questions or having critiques and criticism of foreign policy. It's discussion and debate.
What's frustrating to me, and I hope that people who may have angrily voted against Republicans because of perceived policy failures realize this, is that the DSA and the Democrats do not care about affordability. That issue has gone out the window. Two or three weeks in, we're saying our main concern is reparations and relocation of the people of Palestine. It's like, “What?”
At least there would be some consistency if they said, “We want some of the aid that we gave to Israel returned and distributed to the people here.” But it's not even that. It's basically, “We can't stand the Jewish people occupying the area they've occupied since 1948, so we want them out.” That's your platform.
Ed Vitagliano: I think part of the problem is that these two approaches get conflated. They get mixed up, and it's hard to separate them. There are people who have legitimate concerns about the war in Iran, how that got started, and all those things. There are people who are saying, “I don't want to give aid anymore to Israel.”
Tony: I don't disagree with that. I'm not saying that was my position.
Ed Vitagliano: I'm saying if the DSA were for the American people, if they were genuine in their convictions, they would not be advocating this.
Tony: I just think people are conflating these two things, and the DSA is clearly in favor of the Palestinian cause and thus anti-Israel because they don't think Jews should be there. I think they are motivated by antisemitism.
And I need to be critical here: I think the Trump administration and his team have not helped clarify that.
Ed Vitagliano: Yes, that would have been helpful. They could explain what you just did and bring the Republican Party back together on the issue, but right now it's fractured.
All right, folks, this is what we do on this show and on a lot of the programs on American Family Radio, the talk programs, the preaching programs, and programs about family and finances. When we talk about politics, we try to explore these issues from a Christian perspective. I hope it was helpful to you today. There is more great programming directly ahead on American Family Radio, and, Lord willing, we'll see you here on this show tomorrow.
Ed Vitagliano: And welcome back. Ed Vitagliano sitting in for Tim Wildmon today, joined in studio by Wesley Wildmon and Tony. And now Steve joins us, living up to his name. Very nice paisley shirt again.
Steve: When you go paisley, you go full paisley.
Ed Vitagliano: I go full paisley.
Steve: You have some nice shirts.
Ed Vitagliano: See, you have a life. Represent. When the season rolls around, do you have any Christmas paisley shirts or anything?
Steve: I don't know. Christmas colors, maybe.
Ed Vitagliano: That's something to do this afternoon.
Steve: Yeah, I'm going to be shopping for shirts.
Ed Vitagliano: Heather, I'm spending some money, but it'll be well worth it. You wait until you can check me out in my new Christmas paisley shirt.
Steve: All right.
Ed Vitagliano: All right, take it off.
Steve: So I know you guys were talking a little bit about the outages and the meeting up on the Hill. Did you guys tease the clip that I was going to bring in?
Ed Vitagliano: We forgot.
Wesley Wildmon: I don't think we did.
Ed Vitagliano: We did forget.
Steve: I hope I'm coming in.
Ed Vitagliano: Why don't you tease it real quick?
Steve: I think they're just coming because of you, Steve.
Ed Vitagliano: You know what? I need them for probably fifteen minutes.
Steve: Well, I want it to look.
Ed Vitagliano: Yeah, go ahead.
Steve: Okay, so we've been talking about all these outages, the FAA ground stop, Microsoft 365, and all these things. They are having a meeting on the Hill right now with Secretary of Homeland Security Markwayne Mullin, the deputy director of the FBI, and the national security adviser. A Republican from Mississippi, Michael Guest, asked specifically about this and whether they're related, and I want you to hear the replies.
You are going to hear first from Markwayne Mullin, then Jay Clayton, who's the director of national intelligence, and then Christopher Wray, who's the deputy director of the FBI. They are all addressing whether or not all these outages are related to cyberattacks.
Michael Guest: Are the outages that we've seen within the last week, do you have information that they are related? And then also, a second follow-up question: Do you believe that they pose any threat to national security?
Markwayne Mullin: We have seen nothing that would indicate that they are related. As the secretary said, after doing a great deal of due diligence inside the government, with the private sector, and with people in the private sector whose job is cybersecurity, we have no indication that these incidents are related or that they're coordinated. We also have no indication that they're the result of a state-sponsored malign actor.
Jay Clayton: We have our cyber action teams. Our cyber experts have been engaged from day one on all of these incidents very heavily. And I want to be very clear on this: At this point, there is no indication whatsoever that there is any cyber malicious activity related to any of those incidents.
Steve: All right, so today Apple is reporting that an internal software update caused disruption to Apple Music, the App Store, AppleCare, and iCloud. Marriott International users are having problems signing into their website. So we're being told that's all coincidental. Do you believe that?
Ed Vitagliano: Well, I talked to Bob again, who is kind of our expert on AI, and he says this has all the trappings of what happened a couple of months ago when Anthropic's AI bot got out of its sandbox, which is basically a place where it's supposed to be isolated from the outside world. It escaped and did a whole bunch of malicious damage. He says something similar seems to him to be going on here.
Steve: Are you saying that this could be what we're seeing, kind of an Anthropic-out-of-the-sandbox-type attack?
Ed Vitagliano: Yeah, I think it's all suggesting that. Now, I don't know that that's true, and I don't know that Anthropic, OpenAI, Google, or anyone else, including DeepSeek, is involved in any of this. But it seems as if, months after the initial acknowledgment that things got out of control, we began to see that it was really out of control, far worse than they initially established.
With a flurry of activity like we're seeing today, this suggests that, yes, indeed, things are a bit out of control, and that we need to be very concerned and very active in doing whatever we can to get it back in the sandbox.
Ed Vitagliano: All right, Tony, I want you to explain a thing or two here, because when I hear what Steve was saying and what Bob McGinnis is saying, Steve does a lot of stories on this, and you get to the place where you use the lingo of companies and platforms that are using AI, but most of us don't use that kind of language.
When both Steve and Bob McGinnis were talking about Anthropic escaping out of the sandbox, the AI agent, I think of Planet of the Apes, the first of the trilogy from a few years ago. These intelligent apes got out. They escaped from captivity. That's what it sounds like. What does it mean to say an Anthropic AI agent escaped from the sandbox? Bob McGinnis is saying some of what has happened here recently may be another example of that. Please explain it as if our listeners and I are listening.
Tony: Or as if Wesley is listening.
Ed Vitagliano: Or if Wesley is listening, explain it so that we understand what Bob is talking about.
Tony: I don't know. I'll try to come up with a metaphor.
Ed Vitagliano: You already gave me the metaphor.
Tony: Okay, sandbox is a metaphor.
Ed Vitagliano: Sandbox is a metaphor. Kids play.
Tony: Anthropic, in order to continue training their AI and improving it, obviously does tests with it. You put it in what you call a sandbox, where you give the AI freedom to do tasks that you normally would not want it to do in the wild or out on the internet.
One of the trainings they were doing with this AI was trying to see how proficient it was at hacking other systems independently, using just prompts and commands. What happened is that Anthropic claimed the AI they were training got out of the sandbox, escaped the gorilla, escaped from the cage, and started tearing things up in the lab.
Ed Vitagliano: Okay, right, and so this AI got out and was on the internet.
Tony: It was on the internet. It was doing real-world problems, causing some issues. They said it didn't do anything detrimental, but I think what is interesting about what Bob McGinnis is saying is this: I'm not saying this is what it is, but these attacks, or not attacks, these crazy coincidences could be caused by AI.
I'm not so sure about wild AIs that are just doing their own thing, but AI could be used by actors. You could just set an AI loose and give it commands and instructions, and it would continue seeking out these systems, seeking out vulnerabilities, poking and prodding. If it finds something, it's going to go in and exploit that weakness.
Some of these outages that we're seeing may indeed be caused by some of these companies rushing to patch vulnerabilities. They are rushing so fast that they are breaking things. But you're just asking a lot of me to believe the government in this instance, that these aren't related at all.
If you have a line of cars, and we all have the same car, and Brent's transmission goes out, that's weird. That stinks for Brent, but we're sorry about Brent. Then Steve calls me the next day and says, “Hey, my transmission went out too.” Maybe we need to look into that. Then I call the next day and say, “My transmission just went out.” At what point do you say, “Maybe there's an issue with the manufacturer”?
You would take that same logical course of action and apply it in this example. But it's like the government, maybe they don't want to cause panic, refuses to say that these issues are connected. Those are critical systems: the FAA, major banks, the NSA, Microsoft. Microsoft was out yesterday, right?
Ed Vitagliano: You're just asking a lot of me to believe you when you say they're not related. Wesley, what do you think? Am I crazy?
Wesley Wildmon: I had two questions for you. Is it possible? No, no, I know it's possible. Do you think that Anthropic recognized a hole that was left intentionally and then escaped out the “don't go out that door” sandbox?
Ed Vitagliano: Back in a couple of schools, you're asking right now. When did this happen?
Wesley Wildmon: The Anthropic thing happened a few months ago, probably back in the summer. But here's my question: Could it have gotten out on purpose, or could it have been allowed out? Were there any consequences?
Tony: I think with the Anthropic situation, you have to sift through what companies will do sometimes. Anthropic is not the only AI company that has said, “Oh, no, our AI is so advanced that it is able to get out and do these things,” but it never actually does anything really, really damaging.
Ed Vitagliano: They would be held liable for it.
Wesley Wildmon: That's what I was getting at. Would it not be beneficial for the company if it does get out, and then they can learn from it?
Tony: Well, I don't think that's one of the things you said.
Wesley Wildmon: Well, there's a bright side. We are at least learning what it can do and how to defend against it.
Tony: But they said it was an accident.
Ed Vitagliano: I will say this: I opened by saying Apple just reported that it disrupted services, and Marriott, on any other given day, unrelated, if Apple services were disrupted and people were having trouble with the Marriott website, no one would notice except for Apple and Marriott customers. It certainly wouldn't come up on this show.
Steve: Right.
Ed Vitagliano: But because of all the things that have happened, now we're just ticking off different things.
Tony: Yeah, I was just trying to tickle the ears of our conspiracy listeners here.
Ed Vitagliano: Well, they're ready.
Steve: Some of them may be in this very room.
Tony: Some of them are in this room.
Ed Vitagliano: One of them.
Tony: I am looking in the mirror to admit that.
Ed Vitagliano: That's why I use the example of the car. So, Steve, if Brent's transmission goes out this year, and a couple of years from now Steve's goes out, and mine a few more years after that goes out, we're not going to think, “Maybe the manufacturer had a problem.” But if they go out the same week, one day after the other, I don't know, it looks kind of coordinated.
And part of the reason, by the way, is that Tony's degree is in computers, and he heads up our IT department. So we're asking him questions for a reason.
Tony: Did he pay for it?
Ed Vitagliano: Did he call us to agree?
Tony: Well, in part, I did. I ate his food and my leftovers when he was still living at home.
Ed Vitagliano: So that's why we're asking Tony questions.
Tony: As these AI platforms and companies are trying to figure out AI, they put it in a sandbox. These AI agents learn, right? These are learning programs, and that's why they're testing their abilities inside that sandbox to see what they're capable of.
The question raised by Bob McGinnis is this: If it's not bad actors and it's not just a coincidence, then it may be possible for these AI agents to have become so adept at learning that they may be doing these kinds of things.
The testing that was going on inside the sandbox was to see if they could figure out how to bring down major U.S. institutions like banking, or somehow shut off the power grid, which is why I think Steve mentioned gain-of-function research. That's what the expectation is with how we got worldwide COVID: They were experimenting to see what happens if you change the DNA of something.
Ed Vitagliano: So is it possible that that's what was happening?
Tony: Yes, and this is what makes me lean toward a coordinated, intentional effort. We don't know what's happening in China. Maybe China's banks are going through the same thing. It's hard to know what's going on with the CCP. Maybe there's an AI agent. Maybe they have a similar allergy.
Ed Vitagliano: With bats?
Tony: With an AI.
Ed Vitagliano: With bats.
Tony: There's no way to know what's going on in Russia or China. Maybe they're experiencing the same issue from rogue AIs. If we got reports of that, I would say, “Okay, maybe rogue AIs are just out there doing this to anybody and everybody.” But we're only hearing about it in the United States and in European countries. And who is opposed to the United States and European countries?
That's what makes me feel like it's coordinated and intentional.
What makes AI so dangerous is this. I'll use another metaphor. I'm giving lots of metaphors to help you understand.
A normal hacker, if I'm going to use this building as an example, is going to try to get into the building. He's going to try the door, then he might go to the other door and check that door, then check windows one at a time. That's what normal hacking looks like.
But with AI coming onto the scene, now you multiply and duplicate the hacker. You multiply him by a thousand and send them all to the building, saying, “Get me in the building.” What are they doing? A thousand people are checking windows and doors. They start digging under the building, doing whatever they can to get in. They might climb on one another's shoulders and go into the roof.
That's what these AIs are capable of. They don't stop. They don't stop learning. They don't get tired. They don't go to sleep.
Ed Vitagliano: There's your answer, or something like that.
Steve: I appreciate that.
Ed Vitagliano: I thought AI existed just to give me a dancing cat playing a fiddle.
Tony: Like an image?
Ed Vitagliano: Yeah, it's doing something. It could be doing something far worse.
Steve: All right. So in this last election, we put a whole new raft of Democratic socialists in office, and they are driving the agenda. They are upping the ante when it comes to antisemitism.
For example, Zohran Mamdani was elected from New York, and Representative Melanie Stansbury was elected from, I believe, Colorado. They are demanding two things. They want the right of return for Palestinians to Israel, or what they would call Palestine, and they are looking for reparations from Israel for land they say was taken in 1948.
Ed Vitagliano: Now, this is a new wrinkle.
Steve: It is. It's upping the antisemitism angle.
The right of return means that whatever ancestral land your parents or grandparents were kicked out of, whether in Jerusalem or Tel Aviv, you have the right to come back and claim that property, kicking the Jews out. You would have a whole lot of Arabs and Jews without fences. It would really mean the end of the Jewish state, because it's kind of hard to coexist like that. They don't get along well without fences.
Then they would have the same problem with reparations that we have here with reparations: Who gets them and who doesn't? But this is a new wrinkle, a new antisemitism that the Democratic Socialists of America are prodding the Democratic Party to take.
Ed Vitagliano: All right, Steve, let me ask you this. I've had some text exchanges with friends and friends of this ministry. They don't like the label of calling everyone an antisemite who might want Israel to explain some things about the war in Gaza or have those kinds of questions.
They say that doesn't make me antisemitic. There are people in conservative circles who believe that Israel dragged us into war with Iran, promising that if we did, there would be regime change. If that is true, they would be critical of the Israeli government without necessarily hating Jews. We could go on and on with these examples.
The Democratic Socialists of America, though, clearly seem to be antisemitic. When you talk about the right of return, in other words, Jews having to give up their property and move out, and paying Arabs for land that was rightfully given back to the Jews in 1948, that's not just questioning the government. That is “from the river to the sea” type stuff. It would result in the murder of many Jewish citizens, which we saw from Hamas in 2023. You're basically placing a whole bunch of terrorists right in the middle of them.
Steve: I think it's fair to say you can question Netanyahu, his role, and the Israeli government without that being antisemitic.
Ed Vitagliano: Do you think that's antisemitic? I'm polling the room.
Steve: No, I don't think it's antisemitic to say, “Can you explain whether you were behind the bombing of Iran? Are you pressuring us to go back again and do it again?” Those are policy issues that are open for discussion, and I don't think that is antisemitic.
Wesley Wildmon: I like the way you set up the question. I would fall into the category of people you've talked to. I also like his answer. I don't have anything to add to this.
Tony: The litmus test for me has always been this: If I would ask the same question about U.S. geopolitical policy, then I should be able to ask the same question about any other country's geopolitical policy. If I had questions or concerns about something the United States was doing globally, the litmus test is that I can ask that about any country, including Israel, Mexico, Canada, or whatever. The list is open to everybody.
But what the Democratic Socialists of America are advocating for, as Steve put it well, is more than just asking questions or having critiques and criticism of foreign policy. It's discussion and debate.
What's frustrating to me, and I hope that people who may have angrily voted against Republicans because of perceived policy failures realize this, is that the DSA and the Democrats do not care about affordability. That issue has gone out the window. Two or three weeks in, we're saying our main concern is reparations and relocation of the people of Palestine. It's like, “What?”
At least there would be some consistency if they said, “We want some of the aid that we gave to Israel returned and distributed to the people here.” But it's not even that. It's basically, “We can't stand the Jewish people occupying the area they've occupied since 1948, so we want them out.” That's your platform.
Ed Vitagliano: I think part of the problem is that these two approaches get conflated. They get mixed up, and it's hard to separate them. There are people who have legitimate concerns about the war in Iran, how that got started, and all those things. There are people who are saying, “I don't want to give aid anymore to Israel.”
Tony: I don't disagree with that. I'm not saying that was my position.
Ed Vitagliano: I'm saying if the DSA were for the American people, if they were genuine in their convictions, they would not be advocating this.
Tony: I just think people are conflating these two things, and the DSA is clearly in favor of the Palestinian cause and thus anti-Israel because they don't think Jews should be there. I think they are motivated by antisemitism.
And I need to be critical here: I think the Trump administration and his team have not helped clarify that.
Ed Vitagliano: Yes, that would have been helpful. They could explain what you just did and bring the Republican Party back together on the issue, but right now it's fractured.
All right, folks, this is what we do on this show and on a lot of the programs on American Family Radio, the talk programs, the preaching programs, and programs about family and finances. When we talk about politics, we try to explore these issues from a Christian perspective. I hope it was helpful to you today. There is more great programming directly ahead on American Family Radio, and, Lord willing, we'll see you here on this show tomorrow.
Ed Vitagliano sits in for Tim Wildmon with Wesley Wildmon, Tony Vitagliano to examine a series of recent outages involving the FAA, banks, Apple services and other major systems. Government officials say there is no evidence the incidents are coordinated or linked to malicious cyber activity, while the hosts discuss whether AI agents or bad actors could be exploiting system vulnerabilities.
The discussion explains how an AI sandbox is intended to contain an agent during testing and considers the risks of AI-powered hacking that can operate continuously across many targets. The hosts stress that the possible connections remain unproven while questioning whether officials have adequately explained the disruptions.
The program also addresses Democratic Socialists’ reported calls for Palestinian right of return and reparations from Israel. The hosts distinguish criticism of Israeli government policy from positions they believe would eliminate Israel as a Jewish state and reflect antisemitism.
Did you know American Family Radio is listener-supported? Your monthly financial support will allow us to continue upholding Godly values through our numerous channels, like American Family Radio, One Million Moms, The Stand magazine, and many more.
Become a Great Commission Partner TODAY to help AFR transform culture by rebuilding the family. When you donate today, you’ll get access to: