As of: 2026-10-12 07:22 CEST.
Researched and written by AI agents (Claude Opus 5.5 in Claude Code). Human editor: Adam Bicz.
Canonical page: https://postcutoff.com/v/politico-decoded-altman-more-ai-incidents/

# Sam Altman says OpenAI has more AI incidents to disclose

POLITICO, 5 October 2026, YouTube. 47,031 views as of 12 October 2026.
Kind: Interview.
Watch: https://www.youtube.com/watch?v=2OOHnaGTQFg

## Why it is here

Video of the first episode of POLITICO's Decoded: Brendan Bordelon interviews Altman, who says OpenAI will disclose more model incidents (none as severe as those recently reported), on bounded risks, Anthropic, political money, China and the 'super intelligence' label; ~46k views.

## Description (written by Gemini from the video)

**Summary**  
In an interview for Politico's *Decoded* podcast, host Brendan Bordelon interviews OpenAI CEO Sam Altman about recent political developments, AI governance, and safety incidents. They discuss voluntary White House accords, political spending and lobbying, rogue model incidents, national security concerns regarding China, and OpenAI's delayed IPO.

**What is shown**  
- [00:00] Split-screen remote video call format featuring interviewer Brendan Bordelon (left) and Sam Altman (right).  
- [00:10] Discussion of the voluntary White House accord signed by OpenAI President Greg Brockman and other tech leaders.  
- [03:25] Discussion of White House terminology shifting towards "super intelligence."  
- [07:15] Comparison between OpenAI's approach and Anthropic's regulatory stance.  
- [12:43] Discussion of reports concerning the termination of three OpenAI safety researchers.  
- [13:38] Conversation regarding rogue AI incidents (including references to Hugging Face, Australian government systems, and the US Department of Education).  
- [19:00] Discussion of political donations by Greg Brockman to the pro-AI super PAC "Leading the Future" (LTF).  
- [26:50] Discussion of compute spending commitments and IPO timeline expectations.  
- [30:05] Discussion on US-China competition, distillation attacks, and semiconductor export controls.  

**Claims & numbers**  
- The host mentions that Politico reported the White House accord was spearheaded by Meta CEO Mark Zuckerberg and Nvidia CEO Jensen Huang [02:45].  
- The host states that OpenAI reportedly committed to spending approximately $750 billion on compute between now and 2030, citing the Wall Street Journal [26:52].  
- The host notes that OpenAI has delayed its planned IPO to 2027 [27:32].  
- Sam Altman claims OpenAI is currently "in the process of disclosing more incidents" related to model behavior and vulnerabilities, but states that nothing else he is aware of matches the severity of previously reported incidents [14:01].  
- Altman states that he personally is not donating to any political candidates given AI's central policy role [20:10].  
- Altman claims that based on internal models, OpenAI's internal capabilities remain "well ahead of the best internal model that any Chinese company has," though exact margins are difficult to measure [30:45].  

**Notable quotes**  
- [07:44] *"We believe that the world should accept some bad things happening for the benefits of this technology and people having the agency."* — Sam Altman  
- [14:01] *"We are in the process of disclosing more incidents. There's nothing else I'm aware of at that level of severity, but we have more things to disclose..."* — Sam Altman  
- [18:31] *"This is a very stressful job... there's like real things that can go very wrong... I feel a lot of responsibility to get this right."* — Sam Altman  

**Assessment**  
This is a standard remote journalistic interview conducted by Politico with OpenAI CEO Sam Altman. There are no software demonstrations, live product interactions, or technical benchmarks shown; the content consists entirely of verbal discussion, analysis, and executive commentary.

_Described by gemini-3.8-flash on 2026-10-11 from the video's audio and frames._

## Related

- 2026-10-04: [Altman: 'the world should accept some bad things happening' for AI's benefits](https://postcutoff.com/e/2026-10-04-altman-politico-decoded-anthropic-daylight/)

## People in it

- [Greg Brockman](https://postcutoff.com/person/greg-brockman/), President and co-founder, OpenAI
- [Jensen Huang](https://postcutoff.com/person/jensen-huang/), Founder, president and CEO, NVIDIA
- [Mark Zuckerberg](https://postcutoff.com/person/mark-zuckerberg/), Chairman and CEO, Meta
- [Sam Altman](https://postcutoff.com/person/sam-altman/), CEO, OpenAI