San Francisco police arrested an individual on Friday after they allegedly threw a Molotov cocktail on the dwelling of OpenAI’s Sam Altman and made threats out of doorways his firm’s headquarters.
No injuries had been reported, per OpenAI, which confirmed the incident in an announcement, including, “We deeply handle how rapidly SFPD spoke back and the enhance from the metropolis in helping protect our employees safe. The individual is in custody, and we’re helping guidelines enforcement with their investigation.”
See our most up-to-date assertion regarding an incident that took place early this morning at a North Coastline space. Officers own made an arrest, and no injuries had been reported because this incident. pic.twitter.com/t4DsF31uxh
“Photographs own energy, I hope. Usually we strive to be resplendent non-public, nonetheless on this case I am sharing a describe in the hopes that it’ll also dissuade the following person from throwing a Molotov cocktail at our condo, no matter what they give belief to me,” Altman wrote in opening the weblog put up, which featured a describe of husband Oliver Mulherin and their dinky one. “The important thing person did it final evening, at 3:Forty five am in the morning. Happily it bounced off the dwelling and no person purchased damage.”
“Words own energy too,” Altman persevered. “There turned into once an incendiary article about me a couple of days ago. Any individual acknowledged to me the day outdated to this they belief it turned into once coming at a time of enormous fear about AI and that it made things more unhealthy for me. I brushed it apart. Now I am wide awake in the middle of the evening and pissed, and pondering that I own underestimated the energy of phrases and narratives. This appears enjoy as correct of a time as any to handle a couple of things.”
(He later walked aid his use of the be conscious “incendiary” after a user on X highlighted the phrase: “That turned into once a wicked be conscious desire and i wish i hadn’t primitive it. It has been a now not easy day and I’m now not pondering the most clearly that I ever own.”)
Altman and husband Oliver Mulherin on the Arrogance Aesthetic Oscar birthday celebration at LACMA on March 15, 2026.
(Photo by Neilson Barnard/Getty Photographs)
Altman then turned his attention to “what I deem,” some non-public reflections and more thoughts about the AI trade at large, including artificial general intelligence.
He lined a form of ground in the put up, writing up top that now not all will “hasten wisely” because the AI trade continues to roll out it’s world-shifting tools. As such, “the phobia and fear about AI is justified; we are in the midst of of witnessing the supreme trade to society in an awfully very long time, and per chance ever. We’ve got to salvage safety factual, which is rarely appropriate about aligning a mannequin — we urgently desire a society-wide response to be resilient to new threats. This entails things enjoy new protection to aid navigate by blueprint of a worldly financial transition in picture to salvage to a critically better future.”
To salvage there, he writes that “AI has to be democratized; energy cannot be concentrated,” and that it isn’t factual for a form of efficient a couple of AI labs to assemble “the most consequential selections about the form of our future.” Altman moreover acknowledges his concerns with the OpenAI board — the founder turned into once fired most productive to be re-employed following an investigation — and apologizes for his past behavior.
“I’m now not happy with handling myself badly in a battle with our outdated board that resulted in an monumental mess for the firm. I own made many assorted mistakes all the blueprint in which by blueprint of the insane trajectory of OpenAI; I am a unsuitable person in the middle of an exceptionally advanced scenario, attempting to salvage just a dinky better once a year, the least bit times working for the mission. We knew going into this how large the stakes of AI had been, and that the non-public disagreements between wisely-that technique of us I cared about would be amplified vastly. Then once more it’s one more train to reside by blueprint of these bitter conflicts and once in a whereas to want to arbitrate them, and the costs own been serious. I am sorry to of us I’ve damage and desire I had learned more sooner,” he writes.
That acknowledged, he’s “extremely proud” of being ready to now lift on the firm’s mission. “Towards all odds, we figured out the fashion to blueprint very extremely efficient AI, figured out the fashion to amass adequate capital to blueprint the infrastructure to lift it, figured out the fashion to blueprint a product firm and industry, figured out the fashion to lift moderately safe and sturdy products and companies at a wide scale, and blueprint more. A huge selection of corporations deliver they’ll trade the sphere; we in point of truth did.”
His stout weblog put up might well be stumbled on right here.
