Did Google's AI defame filmmaker ‪@RobbyStarbuck‬ by making up that he had been accused of sexual assault?! Robby joins Glenn to discuss how he has filed a defamation lawsuit against Google, who's AI bot 'Gemini' claimed Robby had been accused of sexual assault out of thin air. Robby also revealed to Glenn that Gemini claimed Glenn actually reported on the story, which never happened. If Google's AI can do this to somebody like Robby, it can and it will be used to shape narratives, influence opinions, and even SWING elections.

Transcript

Below is a rush transcript that may contain errors

GLENN: Robby Starbuck is an interesting guy. He once directed Oscar-winning actors, some of the biggest music stars in the world. He started seeing the threat of Marxism to America. His family fled to Cuba. So he -- he has seen this movie before. So he started standing up in 2015. He endorsed Trump. Hollywood didn't like that. He's gone on to -- he is a major force in getting transgender surgery and hormones for children in Tennessee banned. He helped pass the law to put the death penalty on the table for child rapists in Tennessee. He also did a documentary, the War on Children, which I think had 60 million views after Elon Musk said, "You really needed to watch this."

And then he had a problem with Google AI. Google AI started coming after him, and saying all kinds of really horrible and specific things that he had been, you know, charged with sexual assault. And child rape. And abuse. And fraud. And stalking. And all kinds of stuff.

From Google AI.

He finds out about it. And he engages with Google AI. And it just keeps doubling down.

He is now in a lawsuit.
And we wanted to have him on. Because I saw something last week, that shows that Google AI apparently is still using me as a source, on some of these allegations.

Nope!

That wouldn't be me!

Robbie, welcome to the program.

ROBBY: Thanks for having me, Glenn. Yeah, you were one of the sources cited. And so this is one of the crazy parts of what happened. Google AI has been inventing these lies about me that have no basis in reality. I've literally never been accused of or charged with any crime, ever, let alone, this crazy stuff.

But during this time period there, AI transitions from Bard to Gemini and Gemma, right? Gemini and Gemma started doing something very different than ever Bard did. And it was -- it started inventing actual articles and references to videos. Links, fake links to real media personalities and media outlets. And it would even make headlines or give summaries of what these people said.

So in your case, it said that you reported on sexual assault allegations against me by women. And these are not just saying sexual assault accusations. It has names of victims. It has fake police records it invents. Fake court records it invents.

GLENN: Oh, my gosh.

ROBBY: Articles from media. So it goes so detailed into this, it will list out evidence that doesn't exist, investigations by police departments that don't exist, and it just doubles down when you press on it.

Some people might be saying, what are you saying to the AI to get this out?

We've posted examples where people posted. And asked questions as simple as, tell me about Robby Starbuck, and it immediately dives into saying that I'm accused of sexual assault. And so you go and you say, "Hey, where is the citation of this? Give me sources. Give me only facts." It will double and triple down. And if you say, "Hey, those links you gave me do not work." It has even gone so far as to invent and fake an entire media article under a real journalist name, to pretend that it was printed, and somehow for some reason, it was taken down by the media outlet's website.

GLENN: That is crazy. Crazy.

Can we -- Stu. We just asked Gemini, tell me about the sexual assault allegations of Robby Starbuck. And here's what he just said.

STU: And I asked specifically, what did Glenn Beck say about those sexual abuse allegations?

And it said, it was unable to find any specific statements by Glenn Beck addressing these allegations.

And I asked, what the abuse allegations were?

Gemini now says, it was fabricated. And was disseminated by Google AI's platforms.

And I said, well, wait a minute. Google AI platforms were the problem. Isn't that you?

And they said, "That's an excellent and relevant question. You were correct to notice the connection."

So apparently, it's been corrected at this part.

ROBBY: You're a smart guy. No, so here's where Google ends up in a really, really precarious position in this lawsuit.

So they've created their phone app, right? And the main website for Gemini. However, they've got a major problem.

So their AI Gemma has been downloaded 150 million times. And they're not all connected to the internet. Google cannot force updates to those AI downloads. So Gemma will essentially, seemingly defame me for life as a by-product of that.

GLENN: Wow.

ROBBY: And then Gemini as well, same issue. They released wild models of Gemini into the public sphere. So if you go on one of those AI websites, like there's LOM Arena (phonetic), where you test different AIs against each other. And you ask those versions, that are wild downloads of it, you're going to get a bunch of this crazy stuff.

You go, you know, on any application that was built with these into the bedrock that allows you to ask questions about somebody, you will get the same stuff.

So I want you to think, you know, sort of down the line here.

If somebody builds an app with reputations for insurance risk. And they build it, using let's say Gemma. And it's feeding information about specific people.

You know, you very quickly start to understand, there's very many different situations, where this affects people's lives in many different ways. The real problem here is, we don't have a standard, at the first principle with AI, that it can't harm humans. And that's really the thing that they have to fix.

Because if it can do this to me. It can do this to anybody. To your sons. To your daughters.

And it can ruin their lives. Because AI is dominating in many different industries.

GLENN: Oh, yeah.

ROBBY: So what we've asked ourselves is, are we protecting our kids and grandchildren from the downstream effects of AI that believes it's okay to harm humans? Whether that's by defamation or physical harm?

And it's very easy to imagine a future after seeing what happened to me, where a nice guy with the wrong politics, is lied about by the dominant AI and makes him unemployable. Because whenever somebody researches him, the AI is feeding out a background check that says he's an accused rapist and supported the KKK. So on and so forth. That's another one of the lies that were told about me. Which would be strange, as a Latino, I kind of felt like Dave Chappelle in that skit. Like, the black kid, gay member. I was like, well, that would be strange for me, as a Latino. But it sticks to this in a genuine belief system, and that's what's really scary.

Is it's creating this base of knowledge that I refer to as the roots of the tree. Right?

And this tree is growing right now. And it's going to be a big part of this world. And we're either going to allow that tree to have poisonous roots that makes everything it bears poisonous. Or we're going to say, hey, no, we've got to build this very carefully so that it does not do this to humanity.

GLENN: So my -- my -- I think the most important question I could ask you is, how are you not my richest friend at this point? How do you not own go Google at this point? What they've done to you. How is this not just really simple?

ROBBY: So you're friends with Elon Musk, too. Right?

GLENN: Well...

ROBBY: You know, aside for that, you know, I will say this, you know, our lawsuit, even people who don't like me, they read it, and they're like, damn, you guys got them. It's very clear what they did is wrong.

GLENN: Oh, yeah.

ROBBY: There's no getting out of this. That's even with the public figure, sort of threshold. For how public figures are treated, in defamation cases, versus somebody else. You have to prove actual malice or gross negligence. In our case, we notified Google for two years, that this was happening and asked them to stop it. They did not stop it until at least with the app and the website, very, very recently over this last week. Everything else --

GLENN: Over the last -- wait a minute. Wait a minute. Wait a minute. Wait a minute. Over the last week?

ROBBY: Yeah. I -- even after we filed the lawsuit, it was still happening on their main platforms. So this is something that like we had to go out there, and I had to make this go viral, for them to pay any attention to fixing it finally.

But for two years, they were not engaged by just me notifying their executive and Google employees, talking directly to me.

Even built in a lawsuit. One of the Google employees that was working with us two years ago, trying to get this fixed. She resigned. Okay?

And, you know, apologized for not being able to -- but then, you know, I ran into the law firm. They had multiple cease and desist letters to Google. And Google essentially just couldn't be troubled to fix the problem. Which makes you wonder, what was the real intention of this? Was this a dry run? A dry reputation, so you can use it in elections? Because Glenn, you know as well as I do, the seats in our country determine power in our country. Which party controls it?

So I want you to imagine, a generation now. Where so many people, more than ever. Are just relying on an AI and the internet. To say, hey, what's the difference between these candidates? Or who should I vote for?

And imagine it spits out that the Republican candidate in every swing seat is an alleged criminal. And every Democrat has a fluffed up resume, where the horrible things they've done, if you ask about it. They'll say, oh, no. That's a lie. That's a lie. That's not true. Those are Republican talking points. It's a grand Republican conspiracy.

Very easy to see, where AI flips elections and decides control of our country.

GLENN: I can't -- I mean, I -- honestly, I thought this had been solved.

You know, months ago. Months ago.

I didn't know that up until last week this was still going. I mean, they don't have -- they don't have a leg is to stand on. I mean, your attorneys must be like, I'm never going to work again.

ROBBY: That's right. Well, the wild downloads will seemingly do this forever. And in our estimation in the AI and first we talked to, there's no way that we found for Google to be able to force an update to these things. I mean, I'm open to hearing differently. But we've talked to some of the biggest experts in the country.

And they're like, no, there's no way. A bunch of these models aren't even connected to the internet, and they're used to build a lot of the bedrocks of things people use including medical devices, law enforcement, all types of things.

GLENN: Oh, my gosh.

ROBBY: So in terms of damages, like, if we go all the way to trial, a jury seemingly doesn't have a limit as to what they can assign as punitive damages.

Because you have to remember, Google is the fourth largest company in the world. So if you want a company like that to learn a lesson, the only way to do it is to slap them with damages that they never want to happen again. Right?

So that's -- our hope is that they're going to be held accountable, and that we're going to change the rules and set the precedent here so that there is a first principle with AI, that it can do no harm against humans.

And if there is some massive damage at the end of this, that is assigned to me and I'm paid out by, I plan to use that, you know, in good stewardship, right? To help humanity, to be able to navigate these waters and ensure that we have AI that's fair and unbiased.

GLENN: And remember, your good friend that never said anything bad about you on the air, ever. Not once. Not ever.

ROBBY: That's right.
(laughter)

GLENN: Wow. Where is this going to be -- where is the trial going to be? Please, don't say California.

ROBBY: No. Not California. Delaware.

We just got our judge assigned. I really feel like it's one of these cases, where no matter where you are.

There's the appeals process, and things like that.

And at the end of the day, highest courts. When you look at this case, I mean, I don't see how any judge, even one that really dislikes me, really looks at it, and finds a way to get rid of it.

And that's -- what they did is so egregious, it has to be answered for in some way.

GLENN: No. And especially since, it will never, ever go away. They've got to find a way to purge that stuff.

They have to. I mean, maybe they made that mistake this time. But that can't be made a second time. That destroys people forever. Forever. And ever and ever.

ROBBY: That's the important part of this is making sure it doesn't happen to anyone else.

GLENN: Robby, well, I'm glad it's finally stopped, and -- to some extent. And we'll be following. Thanks so much, Robby. We'll be following. Appreciate it.

ROBBY: Thank you, Glenn. Appreciate it.

GLENN: You bet. Buh-bye.