Filmmaker Robby Starbuck is suing Meta after its AI platform lied about him committing a crime and insisted that authorities should take his kids away from him – all because he’s fighting against DEI policies. Robby joins Glenn to explain what happened, and it makes Glenn declare this as possibly the most dangerous AI story he has ever heard. “What Meta has done to Robby Starbuck NEEDS to be corrected,” Glenn says, “or they will do it to EVERYONE.”

Transcript

Below is a rush transcript that may contain errors

GLENN: You're going to hear a wild, wild true story, from Robbie Starbuck. When I heard this a few weeks back, I was outraged. Outraged. And I immediately wrote to Robbie and said, whatever you need, I will stand with you on this. Because I don't care, you know, Robbie and I happen to agree on a lot of stuff. And I consider him a friend. But I don't care if he wasn't a friend, or we didn't agree.

What Meta has done to Robbie, needs to be corrected. Or they will do it to everyone.

I will bring Robbie Starbuck in, so he can tell the story.

Hello, Robbie, how are you?

ROBBIE: I'm doing well, thanks, Glenn. I appreciate it.

GLENN: Yeah. Thank you for coming on. And tell the story.

Explain -- explain what has happened to you, because of meta,and their AI.

ROBBIE: Yeah. So for those who are unfamiliar, basically for the last year, my mission, along with my team, has been eliminating the policies at major corporations. And we've done that. With most of the biggest companies on earth.

And with great success.

So at the time, in August of 2024, we were focused on Harley-Davidson. And eventually changed their policies.

But there was one particular Harley-Davidson dealership that was not happy about it. And they were from Bernie Sander's state of Vermont. And they had posted online a screen shot of Meta AI, and it said that I had been at January 6th.

And was, you know, part of the -- you know, events of that day.

And that I had committed a crime.

GLENN: And charged. And charged with a crime.

ROBBIE: Well, we found out about that part later, and we go and we ask Meta AI, because I initially thought, there's no way, one of the biggest companies on earth is inventing a crime that I committed. Right?

And so we go and check it. And lo and behold. Not only, it says that I committed a crime. It says, I was charged and arrested. And it later said, I pled guilty to a charge of disorderly conduct over January 6th. And the weird part, I wasn't in between DC that day.

This was totally invented out of nowhere.

GLENN: Wow.

ROBBIE: And we found out something later, that actually makes it even more concerning. This does not appear to be a hallucination.

GLENN: Yes.

ROBBIE: In the best estimation of myself and the team. And AI experts we've spoken to.

When Meta releases AI models, they release a version it can download and, you know, disconnect from the internet. And it will continue to work. Because it's like their train model. They call 3.0, 3.1, 2, 3.

And so on, and so forth. When you do that, and you try to create the situation, it consistently comes up with the same lies about me. And they range from the January 6 stuff to in general saying that I'm essentially a dangerous person. And actually, had told advertisers not to advertise with my show. Said, don't work with him.

GLENN: What!

ROBBIE: He's officially a dangerous individual. And it even insisted an authority should consider taking my kids from me.

And then --

GLENN: Hang on just a second.

Hold on just a second.

I want the audience to understand. First of all, there's no way, this is a hallucination.

There's no way.

So Meta AI is -- has suggested to authorities that they take Robbie's children away from him because he's so dangerous.

ROBBIE: Yeah, you're going to love this part. Not only they said that, it said that my kids would probably be better off in the care of somebody who is more accepting of transgenderism and DEI. And so, you know, if you think about this, it makes you feel even more targeted, right?

The fact that it's going into that narrative, tells you that, you know, it's very difficult to think this could be a hallucination. It's very, very specific. Consistent with sort of left-wing ideology, right? And so you have to go back to the training. And say, you know, how did it get trained to believe this? And go further with that.

We're very close to a future where people are going to have robotics in their homes. They're trained on AI.

Thousands and thousands of apps have always been trained on meta's AI. And so that's going to last a very wrong time. Right?

The other interesting thing about AI definition is that if you say something that is totally untrue about me, Glenn. And I sue you.

And that goes to court, and they say, Glenn, Robbie -- you know, there is an end point to that, right?

You know, you get justice for it, and so on and so forth.

With Meta's AI though, there is this sort of weird thing that happens, where we can go to court. And the court will say, oh, yeah. They defamed Robbie.

And at the end of that, you still have all these unplugged models in the wild. That they can download it by millions of people.

And you will never put the genie back in that bottle. Because they can't update versions that are disconnected from the internet.

Right? So when you think about the future. You can have robots that are walking around in law enforcement. Thinking that I'm a criminal or wanted individual.

Because there was a hallucination at one point that claimed I was wanted for kidnapping charges. The whole thing is insane. The whole thing is insane, Glenn. So we sued them. Because I think that's the reasonable thing to do. To try to get justice here. Not just for me. But everybody.

This can happen to anybody.

If it can happen to me, it can happen anybody.

If you think about everything from custody cases. To our future elections.

You know, this could rig elections.

GLENN: You know, let me just throw one more thing in.

And then I want to ask you some questions. Meta's AI also claimed that he had appeared on Nick Fuentes' Show. That had spoken at Nick's rallies, and then he supported him. He was engaged in Holocaust denial. He's never denied the Holocaust. He doesn't know Nick Fuentes and wouldn't be showing up at any of his rallies.

This is hard to believe that this is a hallucination. Because I -- I mean, I'm working with AI right now and AI designers. And what a hallucination is, it can't remember. It can't memorize the entire internet. Okay?

It can't memorize all of human history and information. So it's very good at remembering the beginning and the end. And sometimes, it will fill in the middle. But if you ask it, show me the sources on this! It -- it will usually correct itself.

You know, you have to say, that's not an actual source.

And it will usually correct itself. And it doesn't hallucinate the same thing on the same person, all the time.

So for this to continue to get deeper and deeper and deeper with you. And make things up that are worse and worse, shows that there is something in the algorithm, at least this is my feeling. And I could -- I've got an AI expert upstairs in the break. I'll ask him.

But it shows there's something in the algorithm, that is -- is tipping the AI to, you're a bad guy. You're a criminal.

ROBBIE: I mean, that's how it looks to us. And, you know, you have to wonder, you know. When there's no rules.

What kind of effect can an engineer have if they're an activist.

And they decide, I hate this group of people. I want to make them look terrible.

You know, and I -- looked like what you had seen.

But in an election. A lot of people. Especially young people. Are increasingly asking.

What's the difference between this candidate and that candidate?

Trying to figure out who they align with.

And if AI is allowed to say about them.

Like, doing PR for them. And the other side is, we can literally invent crimes, and say, they're wanted for kidnapping or whatever.

You know, you can shift a couple of percentage points on any election on that basis.

And that's something that I think everyone should find unacceptable.

GLENN: So this is -- you want to talk about David versus Goliath? Have you filed suit against them?

And what -- what's the status of all of this?

ROBBIE: Yeah, we have filed suit in Delaware. That's where they're incorporated.

And, you know, I think every legal expert I've seen, out there, has said, dang. This is the best defamation case I've ever seen.

I mean, it's so egregious.

Then Meta came out. We notified them in August 2024 of what was happening.

They didn't do anything about it, to actually fix the problem, until nine months later, when I came out with a video, where millions of people watched it. And it became a press story.

It was all over the media. Then they finally step up and try to do some things, right?

Then, it was just -- you know, I think about a week. Or close to a week after that had already come out.

And put out a statement apologizing, which, again, is like unheard of in the middle of litigation. They putting out a statement essentially saying, yeah. We did it.

You still had to say things that were just total fabrications, total lies about me.

Societies very clear that Meta was struggling to control what they had built.

GLENN: This is terrifying, Robbie. I mean, I talked to a few senators about your case, and they are just as incensed about it as I am. And I'm expecting Meta to be called up on Capitol Hill, regarding this.

But I don't know.

I mean, you know, what -- what -- AI has so much power now on Capitol Hill.

These gigantic companies.

And I'm not sure what anybody is really going to do.

I hope justice actually works for you.

And the system works for you. Because if not, this is the beginning of AI tyranny.

This is -- I mean, if this is allowed to go on. This is the beginning of it.

And it will in four years from now, forget about -- forget about your -- really, your freedom.

Your reputation. Everything will be gone.

If you happen to oppose what the AI program says is the right way to go.

ROBBIE: Yeah. We need serious change in the system.

You know, I think, I'm with you on this.

You know, I pray that the system works, as it's intended to work.

You increasingly, you know, have to wonder, as you watched sort of the rogue judges try to take on the powers of the executive branch.

Just how many holes there are in the justice system.

And how much it is working, as intended.

But I really do pray that it will.

Because I do think there's downstream effects of this.

The people are not seeing very clearly.

GLENN: Oh, my gosh.

ROBBIE: And I've gone down that dystopian rabbit hole, and I've looked at what would mean in 20 years if this is allowed to happen. It's very dangerous.

GLENN: Twenty? Five! It will be five years. It won't be 20 years.

ROBBIE: Yeah.

GLENN: Robbie, how can I help? And how can the audience help you?

ROBBIE: Well, you know, I will say, we just have to consistently keep this story out there.

I think that's really important, that everybody knows what is happening and what is possible.

Because this extends public figures. Like, if you have a custody dispute. And they're deciding reputation scoring on AI.

You can invent all kinds of things. You can lose your kids over this. If you're going and trying to get insurance, you could be denied insurance over being perceived as a reputational risk, because you have the wrong politics, right?

I mean, this is very dangerous. Talk to your lawmakers. Write to them.

Ask them to establish ethical boundaries for AI. Because there has to be some boundaries. There is -- don't get me wrong.

A lot of incredible things AI can be used to. But if you have no boundaries, it's going to be very, very dangerous. And so that would be the first thing.

You know, at this point, I'm handling the legal costs and everything.

It's clear, that things could get so expensive legally. Depending on the strategy, maybe at that point, we go through another round of saying, hey, can people jump in and help?

But, you know, I think we can take this all the way. And I really do believe it's important that we get serious and sincere changes in the marketplace.

GLENN: Count on me, as a serious ally. I don't think I've ever made this offer to anybody. But I want you to give me regular updates. And come on the -- come on the program.

Whenever you have an update. You just call. I'll put you on the air. And you give us an update. Because we will follow this story.

And it's important. I mean, this is -- this is the beginning of the end. Anything you might be afraid of, with AI. If this isn't corrected, it will happen. And it will happen quickly.

So it's got to be corrected.

Robbie, thank you so much.

ROBBIE: Thank you, Glenn. I appreciate it.

GLENN: You bet. Buh-bye.