Molotov attack on Sam Altman's home sparks AI safety debate amid ongoing investigation

In the predawn darkness of a San Francisco Friday, someone threw a Molotov cocktail at Sam Altman's house. The OpenAI chief executive's Russian Hill residence absorbed the impact around 4 a.m., but security personnel moved quickly enough to extinguish the flames before they could spread. No one was hurt. Within an hour, police had arrested a 20-year-old named Daniel Alejandro Moreno-Gama near OpenAI's Mission Bay headquarters, where he had allegedly threatened to burn down the building itself.

Moreno-Gama now faces charges of attempted murder, criminal threats, and possession or manufacture of a destructive device, though prosecutors had not yet filed formal charges at the time of his arrest. What emerged in the days that followed painted a portrait of a young man consumed by a particular fear. Online records showed he had written extensively about artificial intelligence as an existential threat to humanity. On Substack and in various forums, he had argued that AI advancement could lead to human extinction, that the systems being built were misaligned with human interests, and that tech leaders lacked strong moral values. In one post, he wrote that coexistence with advanced AI systems was impossible and that development needed to stop as quickly as possible. He participated in online communities focused on AI activism, though those communities later clarified he had no formal role and that they did not promote violence.

Two days after the Molotov attack, on Sunday morning, another incident unfolded near Altman's residence. Police responded to reports of gunfire in the area. Investigators determined that a vehicle had passed through and someone inside had fired a shot, though they could not confirm whether the house itself was the target. Two people were arrested a few blocks away: Amanda Tom, 25, and Muhamad Tarik Hussein, 23. Police recovered three firearms during a search of their location. Both were held on charges of negligent discharge of a firearm. Authorities have not confirmed whether the two incidents are connected or whether either was specifically aimed at Altman.

Altman himself addressed the attack in an unusual blog post hours after it occurred. He published a photograph of his husband and their one-year-old son, hoping the image might deter anyone else from similar acts. He confirmed the Molotov had struck the house but caused no major damage. He also reflected on the broader climate surrounding artificial intelligence. "The words also have power," he wrote, noting that a recent article about him, which he described as inflammatory, may have contributed to a more dangerous environment. He said he was awake in the middle of the night, angry, realizing he had underestimated the force of words and narratives. He called for the public debate over AI to be de-escalated, asking for less explosive rhetoric and fewer actual explosions against houses.

Yet Altman also acknowledged what he called a "genuine concern" about AI's impact, describing that worry as valid at a moment of profound social change. He defended technological progress nonetheless, arguing it could yield long-term benefits. San Francisco Police Chief Derrick Lew issued a statement saying the department takes firearm crimes extremely seriously and that anyone committing such acts would be arrested and prosecuted with the full weight of the law. OpenAI said it was cooperating with investigators. As of the time of reporting, authorities had not confirmed a clear motive for either attack, and the investigation remained open.