In short
Whether backlash against AI could escalate into more violence, sparked by an alleged arson/attempted murder attack on OpenAI CEO Sam Altman and broader anti-AI radicalization.
Guests
Nick Robbins-Early, New York journalist covering extremism, disinformation, and tech; Sean Fleming, researcher at the University of Nottingham specializing in anti-technology politics.
Key claims
The attacker, 20-year-old Daniel Moreno-Gamma (Spring, Texas), allegedly acted amid a mental health crisis but had posted anti-AI content online. Federal authorities may treat the case as domestic terrorism if political intent is proven. Most anti-AI groups advocate regulation, but a small violent subset raises the risk of “stochastic terrorism” and copycats.
Notable examples
Molotov attack on Altman’s Russian Hill home; attempted break-in at OpenAI with kerosene and an anti-AI manifesto listing CEOs/investors; references to Luigi Mangione; Eliezer Yudkowsky’s “paperclip” existential-risk arguments and past calls for extreme measures (e.g., bombing rogue data centers).
Written by AI. May contain mistakes. Listen to the episode to check what was said.
Chapters
Tap a time to open that second in VOThe Attack on Sam Altman
0:45 to 2:36
Details of the violent attack against OpenAI's CEO and its implications.
“It belonged to Sam Altman, the chief executive of OpenAI.”
Profile of Daniel Moreno-Gamma
2:36 to 4:00
Exploration of the background and motivations of the alleged attacker.
“Is this an isolated incident or a sign of something more dangerous to come?”
The Online Radicalization of Moreno-Gamma
4:00 to 5:24
Examining how Moreno-Gamma's online presence reflects his views on AI.
“He also posted a few times on forums like Pause AI and Stop AI, these sort of activist groups who have condemned his actions and are nonviolent groups but are dedicated towards preventing AI from being developed.”
Legal Proceedings and Charges
5:24 to 7:18
Discussion on the legal implications and charges facing Moreno-Gamma.
“What have we heard from either Moreno Gamma or his family since this incident?”
The Broader Implications of AI Backlash
7:18 to 9:12
Analyzing the potential for violence stemming from AI concerns.
“It was interesting seeing that the FBI and its presser on this said the FBI will not tolerate threats against our nation's innovation leaders, which I thought was a really interesting phrase as well.”
The Philosophy Behind AI Doomers
9:12 to 11:46
Understanding the beliefs of those who fear advanced AI and their proposed solutions.
“He's probably the most famous AI doomer right now.”
Drivers of Political Violence Against Technology
11:46 to 14:03
Examining motivations that could lead to acts of violence against tech leaders.
“But, you know, there's a difference between attacking AI infrastructure and the actual people who are at the head of these companies.”
Critics of AI: Regulation Over Violence
14:03 to 14:41
The majority of AI critics advocate for regulation rather than violence.
“They're proposing regulation, oversight, legal means of addressing these problems.”
Incidents of Violence in the AI Space
14:41 to 15:08
Recent incidents show rising tensions against AI executives and companies.
“Experts in political violence have already raised the alarm about what they call stochastic terrorism.”
The Rise of Non-Violent Anti-AI Movements
15:08 to 15:48
The emergence of various anti-AI groups advocating for regulations.
“He then left behind a sign saying no data centers on his doorstep.”
Show all 16 chapters
Public Opinion: Distrust in AI Technology
15:48 to 17:48
Polls reveal significant public distrust in AI and its impact on society.
“we have started to see this wider non-violent anti-AI movements spring up in response.”
Generational Divide on AI Technology
17:48 to 19:18
Younger generations express skepticism towards AI compared to older ones.
“Do you think this is something that companies seem to be worried about?”
Marketing AI: From Threat to Utility
19:18 to 20:52
AI companies have shifted marketing strategies to address public fear.
“And there has been this sort of tonal shift among the industry recently.”
Grassroots Movements Against AI Development
20:52 to 22:14
Local communities are pushing back against AI data centers and policies.
“And this discontent is obviously filtering through the much wider population.”
AI Industry's Response to Backlash
22:14 to 24:29
Increasing PR efforts and security measures in response to public backlash.
“Coming up, can the AI industry contain the backlash?”
Potential for Radicalization in AI Criticism
24:29 to 25:51
Concerns about copycat violence inspired by extremist acts against AI.
“I think that the AI industry is kind of walking this thin line right now where they want to be seen as benevolent still.”
Transcript
Automatic transcript. May contain errors.0:00Nick Robins-Early:This is The Guardian.
0:09Nick Robins-Early:Today, could the backlash against AI turn violent?
0:25Nick Robins-Early:It happened in the early hours of the morning, a couple of Fridays ago.
0:29Sean Fleming:Surveillance footage shows him in a champion hoodie walking up towards the gate of this mansion, holding a Molotov cocktail in one hand. And then he proceeds to hurl the Molotov cocktail at the house and flee.
0:44Nick Robins-Early:The house was in the exclusive Russian Hill neighbourhood of San Francisco. It belonged to Sam Altman, the chief executive of OpenAI. And the attacker wasn't done there.
0:57Sean Fleming:about an hour and a half later he shows up three miles away at the headquarters of open ai the artificial intelligence company that sam altman is the ceo of and he attempts to break into the entrance with a metal chair and security comes out and confronts him and according to the criminal complaint, he says that he's going to burn down OpenAI and anybody who's inside it. San Francisco police arrest him and find a jug of kerosene and a lighter and what appears to be an anti-AI manifesto in which he has a list of CEOs and investors who he is planning to potentially kill and a letter that includes a mention of Sam Altman and is directly addressed to the CEO and says, if you should survive this attack, then take it as a sign of divine intervention to change your ways.
2:04Nick Robins-Early:The attack marks the first time that the head of a major AI company has faced a threat to his safety. 20-year-old Daniel Moreno-Gamma has now been charged with attempted double homicide, and authorities say they intend to prosecute him to the fullest extent of the law. And for now, there is relief that no one was hurt. But it comes at a time when fears about artificial intelligence and its impact on jobs, power and society are growing. From The Guardian, I'm Annie Kelly. Today in Focus. Is this an isolated incident or a sign of something more dangerous to come?
2:47Nick Robins-Early:Nick Robbins-Early, you are a journalist based in New York. and you report on extremism, disinformation, and tech. Welcome to Today in Focus. Thanks for joining us.
2:57Sean Fleming:Thanks for having me.
2:58Nick Robins-Early:Daniel Moreno-Gamma, the man who allegedly attacked Sam Altman's home. Could you tell us who is Daniel Moreno-Gamma, and what do we know about why he did this?
3:09Sean Fleming:Well, from what we know, he is a 20-year-old who grew up in Spring, Texas, which is a suburb of Houston. And until recently, he appeared to be just a college student who was attending community college. She worked at a restaurant. And he had, before this attack, been leaving quite a large digital footprint, posting anti-artificial intelligence sentiment. In one post, he called for Luigi-ing some CEOs in a reference to Luigi Mangione, the suspect who is accused of killing the United Healthcare CEO. He also went by the username Butlerian Jihadist, which is a reference to the Dune series of sci-fi novels in which there's a human uprising against hyper-intelligent machines.
4:00Sean Fleming:He also posted a few times on forums like Pause AI and Stop AI, these sort of activist groups who have condemned his actions and are nonviolent groups but are dedicated towards preventing AI from being developed. And Moreno Gamma's online writings actually caught the eye of this podcast that was looking to interview people who were radically opposed to AI. And they interviewed him and they've released sections of that interview. All right, well, let's just start here. I'm going to call you Discord Dan. Are you okay with that? Yeah, that's great. They ask him about his posts, and they ask him about whether he thinks violence is warranted against CEOs and against the AI industry.
4:43Sean Fleming:And he kind of balks at that, and he says, no, that was just online posturing. And they even ask him directly, do you think it would be good if someone were to kill Sam Altman? So you don't really think it would be wise for someone to, let's say, kill Sam Altman?
5:03Um, no.
5:04Sean Fleming:He hesitates for a second and then he says, no, that would be bad. It wouldn't do anything. It's just not worth it. I would normally only advocate for violence as the absolute final, I don't want to say final solution, but, you know, the final, you know, final.
5:23Nick Robins-Early:Okay, you get what I'm saying, okay? What have we heard from either Moreno Gamma or his family since this incident?
5:31Sean Fleming:Moreno Gamma himself hasn't said anything. His parents have issued a statement. They believe he had been suffering some mental health issues. He had no criminal record prior to this. They described him as a caring, loving person who would not want to do this while he was in his right mind.
5:51Nick Robins-Early:And tell us, what's he been charged with and what have the federal authorities said so far about this incident?
5:55Sean Fleming:He is being charged with a range of crimes from state and federal authorities, including attempted arson and attempted murder. It seems like a mix of things. There's what the authorities say that has happened, which is that he has driven from Texas to plan this attack. And he's walking up with a very deliberate purpose to burn down Altman's house and burn down OpenAI. Federal authorities have been quite harsh on this. They have said that if it's proven that he intended to carry out this for political reasons, they would treat it as domestic terrorism. They're going to pursue this with the full extent of the law.
6:34Sean Fleming:He is currently set for an arraignment hearing on the 5th of May. What will happen there is that there will be formally charges introduced. And then things will progress from there and we'll see what kind of defense his attorney launches. So far, they have said that he was suffering a mental health crisis. He did not intend to create harm. What's becoming clear is that Daniel's actions appear to have been driven by an acute mental health crisis. And that these charges are quite extreme for what the public defender argues should be more classified as a property crime or as something below attempted murder.
7:18Sean Fleming:This case is clearly overcharged. This case is a property crime at best.
7:27Nick Robins-Early:It was interesting seeing that the FBI and its presser on this said the FBI will not tolerate threats against our nation's innovation leaders, which I thought was a really interesting phrase as well.
7:38Sean Fleming:The FBI is committed to protecting our nation's innovation leaders and the companies that drive our economy forward.
7:46Nick Robins-Early:What about Sam Altman himself? What's he said since this attack on his own?
7:50Sean Fleming:So he posted a fairly lengthy blog after this incident. The initial thing that he posted in that was a photo of his family. Late Friday, Sam Altman posted this image of his family saying, I'm sharing a photo in the hopes that it might dissuade the next person from throwing a Molotov cocktail at our house. He included a lot of other defenses of AI. He also referenced this New Yorker magazine profile, which came out recently, that was a fairly unflattering portrait of him.
8:26Nick Robins-Early:Sean Fleming, you're a researcher at the University of Nottingham and you specialise in anti-technology politics. Were you surprised when you saw this attack?
8:36Sean Fleming:I wasn't surprised at all because I've been predicting that the backlash against technology would become increasingly violent for a few years now.
8:45Nick Robins-Early:And we should say that Moreno Gama's family say that this has nothing to do about his views on AI and that he was suffering from a really severe mental health crisis at the time. But what do we know about his beliefs in general?
8:58Sean Fleming:From his Substack posts, we can see that Daniel Moreno Gama was influenced by Eliezer Yudkowsky. I can't exactly remember the first video I came across, but probably it was Yudkowsky, I imagine. He's probably the most famous AI doomer right now. He wrote a book last year, a bestseller, in fact, titled If Anyone Builds It, Everyone Dies. I am worried about the AI that is smarter than us. I'm worried about the AI that is good enough at AI research to build the AI that builds the AI that is smarter than us and kills everyone. The central argument of these AI doomers is that even a small misalignment between the goals of a super intelligent AI and the goals of humanity could be catastrophic.
9:45Sean Fleming:Even an artificial intelligence with a relatively benign goal. The classic example is an AI that's programmed to maximize the production of paper clips. Well, if it's intelligent enough, it will prevent anyone from thwarting that goal, and it will eventually turn the entire world into paperclips. If we tried to turn the machine off or to modify it, it could manipulate us or even kill us to prevent us from doing that. The AI doesn't hate you, neither does it love you, and you're made of atoms that it can use for something else. That's all there is to it in the end. This is a sort of facetious example, but it gives you an idea of how they're thinking.
10:29Nick Robins-Early:And these AI movements that have grown up in response to that fear, can you tell us a bit about them?
10:35Sean Fleming:So they grew out of an online subculture commonly known as rationalism. And it revolves around some seemingly innocuous themes like eliminating cognitive biases. So these are not romantic opponents of the technological age who long for a primitive society or who wish to go back to an agricultural society or something like that. These are people who are steeped in the discourse of science and technology and who oppose AI as sort of insiders. The vast majority of AI doomers are simply calling for strict regulation of AI. But it's worth noting that some of them have advocated for pretty drastic means.
11:20Sean Fleming:So Eliezer Yudkowsky, in an article in Time a few years ago, argued that rogue data centers should be bombed, that airstrikes should be used to enforce a total ban on the development of AI. And he even suggested that countries should be willing to risk nuclear war rather than let this go any further. Right.
11:46Nick Robins-Early:But, you know, there's a difference between attacking AI infrastructure and the actual people who are at the head of these companies. And as I understand it, the point that that article was making was that countries should be willing to bomb data centers in some rogue state who was violating international bans on dangerous AI development. But Sean, you study anti-technology politics and violent extremism. So when it comes to individuals taking really extreme action, what are some of the drivers? You know, you've just talked about this grand scale, you know, the people worried about the possible end of the world because of paperclips.
12:31Nick Robins-Early:But what about the smaller, more immediate things? You know, are people just worried about how this might affect their daily lives, you know, their jobs? The grievances produced by economic obsolescence might result in sabotage against tech companies, a bit of vandalism or machine smashing.
12:53Sean Fleming:But I don't think economic obsolescence is going to drive a new wave of terrorism. It's concern about the existential risks on the part of extremely alienated people who travel in very strange echo chambers online. I think that's where the real threat of political violence comes from. And these ideologies that we've been talking about that are rationalist and scientific appeal to people with both the know-how and the security clearance. My big worry is that rogue insiders are going to be the ones who turn against the system and do a lot of damage.
13:31Nick Robins-Early:There have been quite swift moves by the industry and politicians already to blame the AI doomers or at least kind of point some finger of responsibility at them for things like this attack on Sam Altman's property. Do you think that they do bear some responsibility for that? I think the most extreme of them certainly do bear some responsibility.
13:55Sean Fleming:But I don't think that it's fair to blame the pause AI and stop AI movements as a whole. And it bears repeating that most critics of artificial intelligence are not proposing anything particularly drastic. They're proposing regulation, oversight, legal means of addressing these problems. So the vast majority of people who travel in these message boards are not contemplating assassinations of corporate executives, and they're banning people who even mention such things. So I think we should be careful not to tar them all with the same brush. The anti-AI community is large, and only a small segment of it condones violence.
14:37Sean Fleming:But even that small segment is sizable. Experts in political violence have already raised the alarm about what they call stochastic terrorism. Individuals with different causes who carry out targeted assassinations in an attempt to be heroes. I don't think this is the last attempt on the life of an AI executive we'll see in this decade.
15:07Nick Robins-Early:Nick so so far we've been talking about this one act of violence that was perpetrated against Sam Altman last weekend and you know it's worth saying that there have been a few other incidents as well open AI staff were forced into lockdown after threats in November 2025 and in April this year an Indiana politician reported someone shooting at his front door. He then left behind a sign saying no data centers on his doorstep.
Read the full transcript
15:36Sean Fleming:That handwritten note tucked under a doormat covered in broken glass
15:40Nick Robins-Early:where City County Councilman Ron Gibson says he and his son awoke to gunfire. And as this technology has been advancing, we have started to see this wider non-violent anti-AI movements spring up in response. What do these groups look like at the moment?
15:58Sean Fleming:Well, I think there's different sections of it. There's groups that are more involved in calling for regulation and policies to rein in, though not entirely stop the advancement of AI. And then there are more extreme groups and extreme individuals who have viewed this as an existential threat and are saying that this should be stopped by any means necessary. It's a technology that is extremely divisive. It has widespread disapproval within the United States. And a lot of people just really do not trust these companies or these individuals who are running it. There was an NBC poll last month where I believe maybe only 26 % of people in the United States had a positive opinion of AI.
16:47Sean Fleming:In the list of major figures or entities, AI ranks below Donald Trump. It ranks below the Republican Party. It actually ranks a couple points above the Democrats, but pretty much anything else you can think of, it's less popular than. It's not a technology that people view as having positive outcomes. And that's not just on the workforce. People view it as something that will make people less able to connect with friends and society. They view it as something that will make people less creative. There aren't a lot of people outside of the industry and outside of specifically male researchers, studies have shown, that think that this is going to have net positive effects on society.
17:38Nick Robins-Early:And it was interesting looking at these polls, this anti-AI sentiment is especially prevalent among young people, right? The kind of people that you would think maybe would be keen to adopt new technologies, Gen Z. They really don't like AI, do they? Do you think this is something that companies seem to be worried about?
17:55Sean Fleming:You know, any technology would love to have a younger user base, because it's a user base that you can count on for a much longer period of time. The perception that the user base of these is a ton of teenagers generating everything is not really the case. Certainly, there's a lot of people using it for school, for writing essays, for all these things. But I think with a little bit more digital literacy, maybe younger generations are a little bit more sceptical of this technology in the first place. So it's been interesting to seeing in polling that the people most enamoured with it are actually older generations, while younger people are maybe worrying about what it's going to do to their future or worrying about just the impacts that it's having on their society as they see it right now.
18:42Nick Robins-Early:Do you think that all of these very kind of catastrophic views that they're expanding about the future of their own technology is contributing to this.
18:51Sean Fleming:Yeah, I mean, I think that that's definitely part of it. I think that when you have a sales pitch that is, in some parts, our technology could destroy the world if it's not being reined in, our technology is going to erase all these jobs and transform the labor market. I think AI will probably, like most likely sort of lead to the end of the world. But in the meantime, there will be great companies created with serious machine learning. It's not an appealing sales pitch for a lot of people. And I think that these companies are beginning to realize this. And there has been this sort of tonal shift among the industry recently.
19:25Sean Fleming:They've been putting out policy papers saying, actually, what if we were to have a redistribution of the wealth that AI creates back to citizens? And what if there was a four-day work week that the average worker could have as a result of AI? And I don't think it's a coincidence that these sort of talking points are coming as there is increasing calls for regulation. I think the industry is kind of realizing that they need to shift tack on what their messaging is because of the same issues that we talked about.
19:55Nick Robins-Early:But where did this messaging come from? Who is it directed towards? Because it kind of feels like the cat's out the bag a little bit. I mean, people have taken that narrative up and are believing it, right? are believing that their jobs are going to be gone within four or five years. And why was it that this was their kind of pitch?
20:15Sean Fleming:Broadcasting your technology as this world changing or world ending thing is valuable from a marketing perspective when you have a technology that's still emerging and needs people to buy in and needs people to take it seriously. You know, it's a lot easier to raise billions of dollars when you say that your technology can completely upend the world than it is saying, oh, well, this is, you know, pretty good at writing emails. It's useful to a point. But then after that, when there's been this widespread adoption, you kind of have to switch marketing strategy because all of a sudden the things that you've been saying are not really as appealing for the average person who's using this.
20:53Nick Robins-Early:And this discontent is obviously filtering through the much wider population. How is that unhappiness playing out in local communities who might be worried about their electricity bills and data centers being built. Could you tell us a bit about what we're seeing happen, how that's manifesting?
21:11Sean Fleming:Yeah, I think that there's a number of different grassroots movements sort of building around this. There's people advocating their lawmakers for more stringent policies around AI. There is certain candidates right now who are eyeing midterm elections who are running on platforms that heavily focus on saying, look, look, we're going to pass these bills that reign in AI or offer protections. And at a community level, we've seen data centers try to go up in places around the country and face quite a bit of opposition locally. There's people who are saying, we don't really want this in our community.
21:48Sean Fleming:We're worried about what this might do to our electricity bill. We're worried about this technology in general. And with data centers, they create a lot of jobs initially in terms of construction. but they don't really take that many people to run. So they're not really like big boons to the community in terms of job creation. You can run a massive, massive data center with a few dozen people.
22:14Nick Robins-Early:Coming up, can the AI industry contain the backlash?
22:32Nick Robins-Early:How do you think the AI industry will react to not only this growing discontent, but the potential for, on the very radical end of that, more violence against them or their companies?
22:48Sean Fleming:I think that we're going to see a lot more public relations efforts from AI companies and a lot more attempts to humanize both their leaders and the companies themselves. And then on the other hand, we're just going to see a lot more security go up. And this is something that's been happening with executives and different business leaders for a little while now, certainly since the UnitedHealthcare shooting. But there is a paranoia around who might be out to get them. There is, I think, a growing fear around what could happen when you see extremist acts that target specific business leaders or politicians.
23:30Sean Fleming:I think the people who are within that class begin to look around and think, OK, well, if this happened to them, it could happen to me.
23:38Nick Robins-Early:Do you think AI companies will see an opportunity in blaming these kind of grassroots movements for people's unhappiness with their technology as a way of shutting them down?
23:48Sean Fleming:I think that there's always a risk in any kind of political movement where when there's either politicians or companies or special interests that don't want to face some sort of more activist section of the public, that they would potentially demonize those sections of the public. Like, it's possible that I think, like, the AI industry could say these are more extreme factions who are calling for this and they should not be paid attention to.
24:19Nick Robins-Early:Or shut down in terms of, you know, criminalizing them, which is obviously, you know, not the first time that that's happened in terms of trying to do that to protest groups.
24:29Sean Fleming:I think that the AI industry is kind of walking this thin line right now where they want to be seen as benevolent still. And I think any sort of more draconian actions could be viewed as potentially risky and cause more backlash than they would intend. But, you know, that hasn't stopped AI companies from really heavily lobbying for more lax political action towards them. In some cases in California, OpenAI subpoenaed a number of groups who it believed were potentially working against it somehow and in league with Elon Musk. And they subpoenaed these non-profits saying, we want your communications and we believe that you've been, as part of an influence campaign, taking action against us.
25:15Sean Fleming:So that kind of speaks to, I think, the level of lengths they are willing to go in this, even if they might not be saying, look, we need to criminalize these groups. They're certainly trying to undercut them.
25:27Nick Robins-Early:And finally, you know, there has been suggestions that this incident at Sam Altman's home might lead to other kind of copycat anti-AI violence in the coming months or years. Where do you see all of this heading?
25:44Sean Fleming:Whenever you look at radicalism or extremist attacks, if that is what this ends up being, I think that there's always the risk of copycats. And I think it's notable that in Moreno Gamma's post, he referenced Luigi Mangione. And so there's already a sort of knock-on inspiration that happens whenever there is these types of violence or attempted violence. Certainly with regards to his purported manifesto, you don't really write a manifesto unless you want to inspire others. There is certainly an attempt to encourage this when you have something that is as divisive as AI and is as prevalent. And when you have a level of online radicalization that is just permeating society writ large these days, it's going to look for targets.
26:37Sean Fleming:And I think the AI industry, yeah, could very well be one of those targets in the future.
26:43Nick Robins-Early:Nick, thank you so much for joining us today.
26:46Sean Fleming:Thanks for having me.
26:49Nick Robins-Early:and that's it for today my thanks to nick robbins early and sean fleming and you can read all of nick's reporting at theguardian.com this episode was produced by ned carter miles and presented by me annie kelly the sound producer was ross burns and the executive producer was sammy kent and we will be back later on today with the latest
27:19Sean Fleming:This is The Guardian.
