Suspect in Sam Altman Arson Attack Allegedly Had List of AI CEOs, Prosecutors Say

At approximately 4 a.m. on Friday, April 10, 2026, a 20-year-old man from Spring, Texas, allegedly threw a Molotov cocktail at the gate of OpenAI CEO Sam Altman's San Francisco residence, setting the exterior alight [1][2]. No one was injured. Within the hour, the same man was captured on surveillance footage smashing the glass doors of OpenAI's San Francisco headquarters with a chair, telling witnesses he wanted "to burn it down and kill anyone inside" [3][4]. He was arrested on-site, carrying kerosene, a lighter, and a three-part document that prosecutors say amounts to a manifesto calling for the murder of AI industry leaders [5].

Daniel Moreno-Gama now faces two counts of attempted murder, attempted arson, and nine additional state charges, along with federal charges for attempted destruction of property by means of explosives and possession of an unregistered firearm [2][6]. If convicted on all state counts, he faces 19 years to life; the federal explosives charge alone carries a mandatory minimum of five years and up to 20 years in prison [6]. And prosecutors have signaled that this may not be the end of the charges.

"If the evidence shows that Mr. Moreno-Gama executed these attacks to change public policy or to coerce government or other officials, we will treat this as an act of domestic terrorism, and together with our partners, prosecute him to the fullest extent of the law," U.S. Attorney Craig Missakian said [2][6].

The Manifesto and the List

The document recovered from Moreno-Gama is structured in three sections, according to the Department of Justice [5][7].

The first, titled "Your Last Warning," advocates for the killing of CEOs of AI companies and their investors. It includes what prosecutors describe as the names and addresses of "apparent board members and chief executive officers of AI companies and investors" [3][5]. Law enforcement has not publicly disclosed which specific individuals were named, in order to protect those on the list [4].

The second section, "Some more words on the matter of our impending extinction," lays out Moreno-Gama's belief that artificial intelligence poses an existential threat to humanity [3][7].

The third section is a letter addressed directly to Altman: "If by some miracle you live, then I would take this as a sign from the divine to redeem yourself" [7].

Moreno-Gama also emailed a version of the document to representatives at his former college, Lone Star College in Texas, on the day of the attacks [7]. The FBI subsequently raided a home in The Woodlands, Texas, connected to the suspect [8][9].

Who Is Daniel Moreno-Gama?

Moreno-Gama, 20, was a student at Lone Star College in Texas [7][8]. His online footprint suggests a sustained preoccupation with AI's dangers. He used the Instagram handle "butlerian_jihadist" — a reference to the Butlerian Jihad in Frank Herbert's Dune universe, the fictional war in which humanity destroyed all thinking machines [7]. He maintained a Substack account with writings arguing that AI may lead to human extinction [7].

He was also a member of the public Discord server for PauseAI, a legitimate advocacy organization that calls for halting the training of the most powerful AI systems. According to PauseAI, Moreno-Gama joined approximately two years ago, posted 34 messages, and had no other involvement with the organization. Moderators flagged one of his messages as ambiguous, which earned him a warning [10]. After the attacks became public, PauseAI banned his account and issued a statement: "We wish safety and peace to Sam Altman, his family, and everyone affected. Violence against anyone is antithetical to everything we stand for" [10].

The SF Standard reported that Moreno-Gama had been in San Francisco for several days before the attack, carrying a gun along with his manifesto and target list [7]. Prosecutors allege he traveled to California specifically to kill Altman [5].

A Second Incident — and Copycat Fears

Two days after the Molotov cocktail attack, on April 12, police arrested two additional individuals — Amanda Tom, 25, and Muhamad Tarik Hussein, 23 — after a firearm was discharged near Altman's property [11]. An OpenAI spokesperson said the incident appeared unrelated and there was no indication Altman's home was specifically targeted [11]. The proximity in time, however, amplified already heightened security concerns.

Federal authorities have expressed alarm about potential copycat attacks. The alleged manifesto's inclusion of other executives' names and addresses represents what security analysts consider a direct escalation [12]. Fox News reported that the attack "has signaled a volatile new era of anti-tech extremism, sparking fears among federal authorities that this could ignite a wave of copycat strikes against high-profile executives" [12].

San Francisco District Attorney Brooke Jenkins announced her office would "aggressively pursue charges," citing "a dangerous escalation in the debate surrounding artificial intelligence" [12].

The Threat Landscape: Executive Targeting Is Accelerating

The attack on Altman did not occur in a vacuum. According to ASIS International, the leading security industry association, incidents targeting senior executives doubled from 2024 to 2025 [13]. Physical threats — including assaults, stalking, kidnapping attempts, and protest-related actions — accounted for 85% of documented incidents, with the technology and financial sectors facing the highest rates [13].

Source: ASIS International

Data as of Feb 1, 2026CSV

The trend parallels a broader shift in public attitudes toward corporate leaders. After the killing of UnitedHealthcare CEO Brian Thompson by Luigi Mangione in December 2024, an Emerson College poll found that 41% of respondents aged 18 to 29 said the killing was "acceptable" or "somewhat acceptable" [13][14]. Mangione's case generated significant online sympathy, with many social media users framing him as a folk hero protesting corporate greed [14]. Prosecutors in that case warned that Mangione had "inspired others to embrace violence over reasoned debate" [15].

The Soufan Center, a nonprofit focused on global security, warned in a November 2025 report that growing public opposition to AI — driven by labor displacement concerns, environmental impacts of data centers, and broader anti-corporate sentiment — "has the potential to turn violent," with online commentary increasingly citing specific CEOs by name and referencing their wealth [13].

How Much Do Companies Spend to Protect Their CEOs?

The attack has renewed scrutiny of executive protection budgets across the technology sector. According to SEC proxy filings, the spending varies dramatically [16][17]:

• Meta spent $23.4 million on personal and travel security for Mark Zuckerberg and his family in 2023
• Alphabet spent $6.8 million to protect Sundar Pichai in 2024
• Nvidia spent $3.5 million on security for Jensen Huang in 2024, up from $2.2 million the prior year
• Apple spent $1.4 million to protect Tim Cook in 2024

Source: SEC Proxy Filings / Fortune

Data as of Aug 16, 2025CSV

OpenAI's security spending for Altman has not been publicly disclosed, as the company is not yet publicly traded and does not file proxy statements. The median corporate spending on executive security across all industries reached $95,000 in 2024, a 120% rise over three years [17]. But tech companies face above-average risk: nearly 70% report heightened threats, compared to a 45% average across all industries [16].

Hours after the attack, Altman posted a photo of his husband and their toddler in a blog post, writing: "Normally we try to be pretty private, but in this case I am sharing a photo in the hopes that it might dissuade the next person from throwing a Molotov cocktail at our house" [12].

The Legal Path: Attempted Murder, Arson, and the Terrorism Question

Moreno-Gama faces charges at both the state and federal level. At the state level, the San Francisco District Attorney's office has filed two counts of attempted murder — one for Altman and one for the security guard present at the residence — along with attempted arson, possession of a destructive device, and attempted criminal threats [2][6].

Federal charges include attempted damage and destruction of property by means of explosives, carrying a mandatory minimum of five years and up to 20 years in prison, and possession of an unregistered firearm, which carries up to 10 years [6].

The domestic terrorism question remains open. Under federal law, there is no standalone domestic terrorism charge; instead, prosecutors can seek sentencing enhancements by demonstrating that the defendant's actions were intended to intimidate or coerce a civilian population or influence government policy through violence [6]. U.S. Attorney Missakian's public statement suggests prosecutors are actively examining whether the manifesto's content — particularly its advocacy for killing AI executives to halt AI development — meets this threshold [2].

The manifesto itself is expected to serve as central evidence of premeditation. Prosecutors in the criminal complaint describe a defendant who traveled across the country with a prepared target list, arrived in San Francisco days before the attack, and carried out a sequence of planned actions [5][7].

The Rhetoric Question: Who Bears Responsibility?

The case has reignited a debate that has simmered since the Mangione killing: whether public rhetoric about corporate executives contributes to a climate that enables violence.

The Case Against Anti-AI Rhetoric

Some commentators have drawn a line from mainstream AI criticism to the manifesto's contents. Moreno-Gama's writings echo arguments made by credentialed AI researchers and policy advocates — that AI poses an existential risk to humanity, that its development must be halted, and that the executives driving it forward bear personal responsibility for potential catastrophe.

The SF Standard framed the question directly, asking whether AI "has met its Luigi Mangione moment" [18]. The concern is that when academics, journalists, and advocacy groups describe AI executives as architects of potential human extinction, some individuals may conclude that violence against those executives is morally justified.

PauseAI's swift condemnation and disclosure of Moreno-Gama's minimal involvement with the organization reflects an awareness of this dynamic. The group's statement emphasized that its mission is rooted in concern for everyone's safety, "including Sam Altman and his loved ones" [10].

But the line between legitimate criticism and incitement is not straightforward. Moreno-Gama's manifesto draws on real concerns shared by serious researchers. In May 2023, hundreds of AI researchers, including executives at OpenAI and Google DeepMind, signed a statement declaring that "the risk of extinction from AI should be a global priority alongside other societal-scale risks such as pandemics and nuclear war" [19]. The question is whether that language, when amplified across social media and absorbed by a vulnerable individual, can become a blueprint for violence.

The Case Against the AI Industry's Own Narrative

The uncomfortable flip side is that the AI industry itself has been a primary amplifier of existential risk framing. Sam Altman personally signed the extinction-risk statement [19]. OpenAI was founded with the explicit mission of ensuring that artificial general intelligence "benefits all of humanity" — a framing that presupposes the technology is powerful enough to pose civilization-scale consequences.

Critics have argued that this narrative serves a dual commercial purpose: it positions the companies as uniquely important, justifying massive investment and light regulation on the theory that only the builders themselves can manage the danger. A 2023 Nature editorial argued that "talk of artificial intelligence destroying humanity plays into the tech companies' agenda, and hinders effective regulation of the societal harms AI is causing right now" [20].

The Lawfare Institute published an analysis arguing that OpenAI has progressively abandoned its safety commitments, noting that the company dissolved its Superalignment team shortly after key researchers departed, despite having publicly pledged to dedicate 20% of computing resources to safety research [21].

If the very companies building AI have told the public — repeatedly and through their most senior leaders — that their technology could end civilization, it is reasonable to ask whether that messaging contributed to the threat environment their own executives now face. This does not excuse violence, but it raises questions about corporate responsibility in shaping public perception of risk.

Federal Response and the Categorization Problem

How the federal government categorizes and responds to threats against tech executives remains an evolving question. The FBI's Joint Terrorism Task Forces (JTTFs) coordinate interagency responses to politically motivated violence, and a new Joint Mission Center created in April 2026 integrates intelligence and operational support across ten federal agencies [22].

The Department of Homeland Security's 2025 Homeland Threat Assessment identified a heightened risk from domestic violent extremists, noting that social media, encrypted messaging, and generative AI tools have accelerated radicalization while reducing the visibility of threat indicators [22]. A 2024 Government Accountability Office report found that "additional actions" were needed to implement an effective national strategy against domestic terrorism [22].

The challenge for law enforcement is that anti-AI extremism does not fit neatly into existing domestic terrorism categories — it is neither straightforwardly political, religious, nor ethno-nationalist. An individual who targets corporate executives based on beliefs about technological existential risk occupies a novel position in the threat taxonomy. Whether the FBI and its partner agencies develop specific protocols for AI-related radicalization may depend on whether the Moreno-Gama case proves to be an isolated incident or the first in a pattern.

What Comes Next

Moreno-Gama is in federal custody. The FBI's investigation is ongoing, with the Texas raid suggesting prosecutors are seeking additional evidence about his planning, communications, and potential connections to others [8][9]. The specific names on his alleged target list remain sealed, but the individuals involved have presumably been notified.

The case sits at the intersection of multiple escalating pressures: rising public hostility toward corporate leaders, genuine disagreement about AI's societal impacts, a social media ecosystem that can amplify grievances into action, and a security apparatus still adapting to a threat environment that did not exist five years ago.

What is clear is that the attack on Altman's home was not a random act. It was, according to prosecutors, a planned assassination attempt by someone who had written out his reasoning, chosen his targets, gathered their addresses, traveled across the country, and carried weapons and incendiary devices to the scene. The alleged manifesto's call for others to follow suit is what makes the case not just a criminal matter but a potential inflection point in how the country understands — and responds to — the human cost of the AI debate.