SafetyPublished Top storyAll news from that day
Altman doesn't know if Anthropic is safe, and doubts it knows OpenAI's work
In a New York Times DealBook Summit video published 18 November, OpenAI's chief executive said the race dynamic is a bad thing and that more caution, not less, is warranted.
Check our sources · 15 facts from 3 sources
Video: The New York Times on YouTube
Key points
- Asked by Andrew Ross Sorkin whether what Anthropic is doing is safe, Altman says: "I don't know. And I think the they also don't really know what we're doing" (27:42).
- Altman says "the race dynamic in the fog of war I think is a bad thing" (27:47) and "more caution, not less caution, is warranted" (27:50).
- Altman says there "will need to be some shared standards on safety" (24:59) and that "the world needs a clear plan from the companies" (20:46).
What happened
Asked by Andrew Ross Sorkin whether what Anthropic is doing is safe, Sam Altman answers: "Well, this gets to what Greg is saying. I don't know. And I think the they also don't really know what we're doing" (27:42). He ties his reply to what Greg Brockman had said, and adds that Anthropic does not know OpenAI's work either.
He goes on: "I think this is a time where more caution, not less caution, is warranted" (27:50), a remark about the industry as a whole.
This was at the New York Times DealBook Summit on 18 November 2026, the day before Anthropic's listing, with OpenAI president Greg Brockman beside him. The New York Times published the video that day on its own YouTube channel, and Sorkin introduces the two as OpenAI's chief executive and president (0:06).
Altman calls the "race dynamic in the fog of war" "a bad thing" (27:47). He says "there will need to be some shared standards on safety" (24:59) and that "the world needs a clear plan from the companies about how we are going to confront this next phase of models" (20:46).
Sorkin asked whether Satya Nadella is right that alignment will never get there. Altman answers: "I disagree with him on this. Uh I think we will be able to align models" (11:44).
On going public, Fortune reported on 15 September that Altman said "right now would be an ill-advised moment to go public". At the summit he says: "we we plan to be a public company at some point" (13:36).
On rivals, Altman says of a comment he thinks came from Elon Musk about maybe being Labradors: "I don't want to be a Labrador. I don't want my kids to be Labradors" (25:34).
After Sorkin raised Anthropic's use of recursive learning, Brockman says: "I think we're the only company where the founders don't have supermajority control right of the AI companies" (26:15). He does not name Anthropic.
Brockman also says OpenAI is "announcing today" that with AMD it got "a year's worth of full generational improvements in five months just using our models" (9:48).
Underlined sentences link to the facts we checked. Dotted words open a short explainer.
Fact 1 of 15 · from The New York Times (YouTube), 18 Nov 2026
Well, this gets to what Greg is saying. I don't know. And I think the they also don't really know what we're doing.
Fact 8 of 15 · from The New York Times (YouTube), 18 Nov 2026
I think we're the only company where the founders don't have supermajority control right of the AI companies
What it means for you
Our viewAltman says he does not know whether Anthropic's approach is safe, and he thinks Anthropic cannot see OpenAI's work either. He directs the call for more caution at the whole industry, together with shared safety standards and a clear plan from the companies. Those are one executive's stated positions. If you buy from either lab, ask what independent evaluation of its safety work it can show you.
It adds no new facts.
Your reaction
We count reactions per story and day, never who reacted. The counts help us choose what goes in the monthly issue. If you are signed in, your own page shows yours too.
We count reactions per story and day, never who reacted. They help choose what goes in the monthly issue.
Check our sources
Every sentence above is checked against these 3 sources.
1 Why Sam Altman and Greg Brockman Say Humans Must Stay in Control of AI
Open the source archived copy-
Sam Altman says in the video, asked by Andrew Ross Sorkin whether what Anthropic is doing, the way it is doing it, is safe: "I don't know. And I think the they also don't really know what we're doing." (27:42). "The they" is the automatic captions' rendering of what was most likely "they".
citeWell, this gets to what Greg is saying. I don't know. And I think the they also don't really know what we're doing.
-
Sam Altman says in the video: "the race dynamic in the fog of war I think is a bad thing" (27:47), and: "I think this is a time where more caution, not less caution, is warranted" (27:50).
citethe the race dynamic in the fog of war I think is a bad thing. Um, I think this is a time where more caution, not less caution, is warranted.
-
Sam Altman says in the video, asked whether Satya Nadella is right that alignment will never get there (Sorkin's wording, 11:28): "I disagree with him on this. Uh I think we will be able to align models." (11:44)
citeI I disagree with him on this. Uh I think we will be able to align models.
-
Sam Altman says in the video: "I do think that there will need to be some shared standards on safety" (24:59), answering Sorkin's question about whether he could imagine the industry working together under some kind of umbrella system.
citeUh I do think that there will need to be some shared standards on safety
-
Sam Altman says in the video, answering Sorkin on whether the way Elon Musk or Dario Amodei build AI is dangerous, that he saw a comment he thinks was from Elon about maybe being Labradors: "I don't want to be a Labrador. I don't want my kids to be Labradors." (25:34)
citeI saw a recent comment I think it was from Elon about you know maybe we could be Labradors. I don't want to be a Labrador. I don't want my kids to be Labradors.
-
Sam Altman says in the video, when Sorkin asks about his reported comment that this is an ill-advised moment to go public: "we we plan to be a public company at some point" (13:36).
citeI I mean, we we plan to be a public company at some point.
-
Sam Altman says in the video: "I do think the world needs a clear plan from the companies about how we are going to confront this next phase of models" (20:46).
citebut I do think the world needs a clear plan from the companies about how we are going to confront this next phase of models
-
Greg Brockman says in the video: "I think we're the only company where the founders don't have supermajority control right of the AI companies" (26:15). The captions give "right" mid-sentence as spoken.
citeI think we're the only company where the founders don't have supermajority control right of the AI companies
-
Greg Brockman says in the video: "we're announcing today that working with AMD, we were able to get a year's worth of full generational improvements in five months just using our models" (09:48). This note rests on his words only; the claim is his and the video does not show the result.
citewe're announcing today that working with AMD, we were able to get a year's worth of full generational improvements in five months just using our models
-
The video "Why Sam Altman and Greg Brockman Say Humans Must Stay in Control of AI" was published on 18 November 2026 on The New York Times' YouTube channel; its description lists the guests as Sam Altman, co-founder and C.E.O. of OpenAI, and Greg Brockman, co-founder and president of OpenAI, and the interviewer as Andrew Ross Sorkin, and links to the Times' live coverage of the DealBook Summit on 18 November 2026.
citeGuests: Sam Altman; co-founder and C.E.O. of OpenAI | Greg Brockman; co-founder and president of OpenAI. Interviewer: Andrew Ross Sorkin, founder and editor at large of DealBook. Filmed live in 2026 at The New York Times DealBook Summit
-
Sorkin says in the video, opening the conversation with Altman and Brockman, "We have anthropics IPO happening tomorrow" (00:42), the automatic captions' rendering of "Anthropic's IPO".
citeWe have anthropics IPO happening tomorrow.
-
Sorkin says in the video that Sam Altman is the chief executive and Greg Brockman the president of OpenAI, which they founded 11 years ago next month (00:06).
citeSam Alman and Greg Brockman founded OpenAI uh 11 years ago uh next month as it happened. Sam is the chief executive. Greg is the president
-
Sorkin asks in the video whether others are going too fast, saying that at Anthropic it appears they are using recursive learning in a way that may be faster than people anticipated (25:54), and Greg Brockman begins his reply with "I want to add two two things" (26:06), the second of which refers to "what Sam's been saying on shared safety standards" (26:38). Brockman does not name Anthropic in the remark about founders' control.
citeDo you think other guys are others are going too fast? ... you look at what's happening at Anthropic, you look at some of the numbers that have come out and it appears that they really are genuinely using recursive learning in a way that I think may be faster than people had anticipated. I I I want to add two two things ... what's what Sam's been saying on shared safety standards
2 Anthropic, PBC: final prospectus (Form 424B4), dated November 18, 2026
Open the source archived copy-
Anthropic's final prospectus (Form 424B4, dated November 18, 2026) says that Class F common stock, to be held by a limited liability company of which Anthropic's seven co-founders will initially be the sole members (the "Founder LLC"), will be entitled to an aggregate number of votes that equals 50.1% of the total voting power of its outstanding shares of capital stock on matters submitted to stockholders generally, including the election of directors other than the Class T directors.
cite
3 Sam Altman, OpenAI, AI safety, pause (Fortune, 15 September 2026)
Open the source archived copy-
Fortune reported in an interview published on 15 September 2026 that Sam Altman said: "I actually think that, given everything happening with safety, right now would be an ill-advised moment to go public." reported by Fortune
cite
Every sentence above is checked against these 3 sources. Each fact has its own address you can cite.
Source 1 · The New York Times (YouTube) · 18 Nov 2026
Why Sam Altman and Greg Brockman Say Humans Must Stay in Control of AI
Sam Altman says in the video, asked by Andrew Ross Sorkin whether what Anthropic is doing, the way it is doing it, is safe: "I don't know. And I think the they also don't really know what we're doing." (27:42). "The they" is the automatic captions' rendering of what was most likely "they".
Well, this gets to what Greg is saying. I don't know. And I think the they also don't really know what we're doing.
Sam Altman says in the video: "the race dynamic in the fog of war I think is a bad thing" (27:47), and: "I think this is a time where more caution, not less caution, is warranted" (27:50).
the the race dynamic in the fog of war I think is a bad thing. Um, I think this is a time where more caution, not less caution, is warranted.
Sam Altman says in the video, asked whether Satya Nadella is right that alignment will never get there (Sorkin's wording, 11:28): "I disagree with him on this. Uh I think we will be able to align models." (11:44)
I I disagree with him on this. Uh I think we will be able to align models.
Show all 13 factsShow fewer facts
Sam Altman says in the video: "I do think that there will need to be some shared standards on safety" (24:59), answering Sorkin's question about whether he could imagine the industry working together under some kind of umbrella system.
Uh I do think that there will need to be some shared standards on safety
Sam Altman says in the video, answering Sorkin on whether the way Elon Musk or Dario Amodei build AI is dangerous, that he saw a comment he thinks was from Elon about maybe being Labradors: "I don't want to be a Labrador. I don't want my kids to be Labradors." (25:34)
I saw a recent comment I think it was from Elon about you know maybe we could be Labradors. I don't want to be a Labrador. I don't want my kids to be Labradors.
Sam Altman says in the video, when Sorkin asks about his reported comment that this is an ill-advised moment to go public: "we we plan to be a public company at some point" (13:36).
I I mean, we we plan to be a public company at some point.
Sam Altman says in the video: "I do think the world needs a clear plan from the companies about how we are going to confront this next phase of models" (20:46).
but I do think the world needs a clear plan from the companies about how we are going to confront this next phase of models
Greg Brockman says in the video: "I think we're the only company where the founders don't have supermajority control right of the AI companies" (26:15). The captions give "right" mid-sentence as spoken.
I think we're the only company where the founders don't have supermajority control right of the AI companies
Greg Brockman says in the video: "we're announcing today that working with AMD, we were able to get a year's worth of full generational improvements in five months just using our models" (09:48). This note rests on his words only; the claim is his and the video does not show the result.
we're announcing today that working with AMD, we were able to get a year's worth of full generational improvements in five months just using our models
The video "Why Sam Altman and Greg Brockman Say Humans Must Stay in Control of AI" was published on 18 November 2026 on The New York Times' YouTube channel; its description lists the guests as Sam Altman, co-founder and C.E.O. of OpenAI, and Greg Brockman, co-founder and president of OpenAI, and the interviewer as Andrew Ross Sorkin, and links to the Times' live coverage of the DealBook Summit on 18 November 2026.
Guests: Sam Altman; co-founder and C.E.O. of OpenAI | Greg Brockman; co-founder and president of OpenAI. Interviewer: Andrew Ross Sorkin, founder and editor at large of DealBook. Filmed live in 2026 at The New York Times DealBook Summit
Sorkin says in the video, opening the conversation with Altman and Brockman, "We have anthropics IPO happening tomorrow" (00:42), the automatic captions' rendering of "Anthropic's IPO".
We have anthropics IPO happening tomorrow.
Sorkin says in the video that Sam Altman is the chief executive and Greg Brockman the president of OpenAI, which they founded 11 years ago next month (00:06).
Sam Alman and Greg Brockman founded OpenAI uh 11 years ago uh next month as it happened. Sam is the chief executive. Greg is the president
Sorkin asks in the video whether others are going too fast, saying that at Anthropic it appears they are using recursive learning in a way that may be faster than people anticipated (25:54), and Greg Brockman begins his reply with "I want to add two two things" (26:06), the second of which refers to "what Sam's been saying on shared safety standards" (26:38). Brockman does not name Anthropic in the remark about founders' control.
Do you think other guys are others are going too fast? ... you look at what's happening at Anthropic, you look at some of the numbers that have come out and it appears that they really are genuinely using recursive learning in a way that I think may be faster than people had anticipated. I I I want to add two two things ... what's what Sam's been saying on shared safety standards
Source 2 · U.S. Securities and Exchange Commission (EDGAR) · 18 Nov 2026
Anthropic, PBC: final prospectus (Form 424B4), dated November 18, 2026
Anthropic's final prospectus (Form 424B4, dated November 18, 2026) says that Class F common stock, to be held by a limited liability company of which Anthropic's seven co-founders will initially be the sole members (the "Founder LLC"), will be entitled to an aggregate number of votes that equals 50.1% of the total voting power of its outstanding shares of capital stock on matters submitted to stockholders generally, including the election of directors other than the Class T directors.
Source 3 · Fortune · 15 Sep 2026 · second-hand source
Sam Altman, OpenAI, AI safety, pause (Fortune, 15 September 2026)
Fortune reported in an interview published on 15 September 2026 that Sam Altman said: "I actually think that, given everything happening with safety, right now would be an ill-advised moment to go public." reported by Fortune
Topics
The morning email
On the mornings we publish: the three top stories and up to four short ones. Free.
We email you a link to confirm. An issue may include one sponsor, always labelled Sponsored · Advertisement. Our emails count opens and clicks, not who made them. Unsubscribe in one click. What we keep