[0:00:00] I think one of the biggest differences between us and some of the
[0:00:32] Hello, everyone, and welcome to the first episode of Decoded, a new podcast from Politico. Decoded is a daily podcast that breaks down what's new in the world of tech policy, artificial intelligence, and all the people involved in reshaping global technology.
[0:01:00] I'm excited to be joined today by Dakota co-author and host Brendan Bordelon. Hey, Christine. Happy to be here, especially because for our first ever episode, I sat down with OpenAI CEO Sam Altman. We talked about the new White House AI accord, the need for new liability standards for autonomous hacking incidents, the influence of AI industry money in U.S. politics, President Donald Trump's attempted superintelligence rebrand, and so much more.
[0:01:23] Today, we're going to play a part of that conversation and chat about Altman's comments. You can find the full interview on YouTube at youtube.com slash politico and more highlights and analysis from our conversation in the new Decoded newsletter at politico.com slash decoded by politico. So without further ado, here's Brendan's conversation with Sam Altman. Sam, you've had an intense week to say the least. So I really appreciate you taking the time to join Decoded. Thanks for having me.
[0:01:51] Yeah, yeah. You were in Washington recently for the state dinner, and OpenAI President Greg Brockman just signed this voluntary accord on AI safety at the White House. Do you expect this new agreement, this White House-brokered agreement, to change OpenAI safety practices in a meaningful way, or is this essentially just more window dressing? Well, I don't think it's window dressing, but I do want to be clear. We care super deeply about safety and would have done and will do the best thing we can for safety, security, and just sort of like...
[0:02:21] Trust and robustness in this technology, whether or not any voluntary agreement is signed. I think it's great to have the industry do this. And I think clearly we're at a time where people are looking for reassurance from the industry and the developers of this technology. But we were founded on a belief that this technology was going to become as powerful as it has now and more powerful in the future.
[0:02:46] And we wanted to make sure that people get the tremendous benefits of it. And we will do that with or without such an agreement. But I'm happy to have it. You know, how do you respond to those who worry that the public is being shut out of these very important debates on AI safety and that really these policies are being crafted by the president and just a handful of tech CEOs and backrooms at the White House? I do share the spirit of that worry. I really think there are two ways this technology can go.
[0:03:12] Very wrong. I mean, there's many specifics we could talk about, but if we think about the kind of general directions that we should be worried about, one, which has gotten a lot of attention recently, is that there could be a loss of control to AI and that we build something that we don't have sufficiently aligned to human values and that allows people to stay in control of the future forever. And so we need to get safety right. But the other way things can go wrong is too much concentration of power.
[0:03:39] And there's been a lot of fear here where, okay, if we do too much to restrict this technology in the name of safety and we end up with a small number of companies or people or perhaps one country that have too much power over this, that also is a way that this can go horribly wrong because people are shut out of the kind of power and benefits and control and agency over the future they should have.
[0:04:04] Our goal here is to sort of be the pragmatic centrists between these paths because we can see how tempting it is to err in one direction or the other.
[0:04:12] Glad you mentioned the very few people in the room here, because Politico's actually reported that this White House accord that was signed earlier was the brainchild of Meta CEO Mark Zuckerberg and NVIDIA CEO Jensen Huang, with basically minimal input from some of the other folks who signed on. Obviously, Greg was there, but it was unclear to what extent OpenAI influenced this. Were you guys blindsided by this agreement? And are you concerned that, you know, maybe some of the CEOs are back-channeling with the White House on this? No, I think we've been talking with our colleagues in the industry for quite like...
[0:04:40] Many, many months and more intensively the weeks before this. And I think people had like a lot of shared goals here. I would welcome something more, but I think this is like a great start that the industry can get behind. I got to ask about the other news out of the White House this week. There was an executive order officially changing, rebranding artificial intelligence to super intelligence. What do you think of the attempted rebrand here? We've been talking about super intelligence for a few years now.
[0:05:10] It's a different thing, right? I mean, superintelligence is like the machine god, right? Well, I kind of think the artificial intelligence thing happened. I think a lot of people would say AGI happened or is close to happening. I think superintelligence does... It still sort of understates the magnitude of what I think is happening, but I think it's a better word than artificial intelligence. So, yeah, I think like it kind of...
[0:05:33] It matched how we've been talking about it, but I think a lot of people continue to use AI as a phrase, as a word. I was going to ask, are you guys going to change the name to OpenSI? No plans to do that. But we do like, we have been talking about superintelligence a lot, and we will continue to talk about it. Yeah, and obviously I think the president in part is driven by this desire to kind of rebrand AI so that it's less scary to people. And I think maybe artificial sounds bad. I mean, does superintelligence sound obviously less scary to people than artificial intelligence?
[0:06:00] Not to me. So that's why I'm curious whether you think this rebrand will actually have any impact on the public's perception of this technology. Yeah, it's not clear to me that superintelligence is a less scary term. I do think it's a more accurate term. But I think people are going to keep calling it AI. Okay, and we won't look out for the corporate rebrand anytime soon. As a funny little story, a while ago, this was a thing I really wanted to do, but it didn't happen. I wanted to get OpenAI.com, OpenAGI.com, and OpenASI.com.
[0:06:30] And as we moved up the ladder from AI to AGI to superintelligence, we were going to change the URL, but brands are sticky. The futures that might have been. Wow. So turning back to your relationship with the president, you've obviously been pushing for tougher safety rules, kind of a shift a little bit for the company. That seems increasingly at odds with the president's claims that AI safety is a hoax, regulations are a globalist scheme. You've had a good relationship with Trump, but why isn't he listening to you when you're asking for stricter rules?
[0:07:02] Well, I think he really does want people to be able to use AI and rely on it being safe and robust and working the way people expect it to work. He clearly wants the companies to just do this themselves. As I mentioned earlier, we're going to make our own products safe no matter what the rules are and no matter what other companies do. I continue to think that as this technology reaches increasing levels of capability,
[0:07:33] And the world has to contend with what this means for society as a whole. We probably will see more calls for sort of some guardrails, but if I'm wrong about that, and if nothing goes wrong along the way, and, you know, we don't need anything there, then...
[0:07:48] That's fine too. Yeah. And obviously the president is talking to other folks as well. He's got venture capitalists in his orbit that think AI should be regulated in a very light fashion. Existing laws they think can clamp down on any safety concerns. And frankly, a lot of those people thought that your company was on their side on this question. But we've heard from some of these, people call them AI accelerationists recently. And to be frank, they feel betrayed. They feel like OpenAI is basically...
[0:08:12] Copying Anthropic and its CEO and its push for very aggressive regulations on AI. Where do you specifically diverge from Anthropic and its pro-regulation CEO, Dario, when it comes to new AI rules? Where's the daylight? I think there's a lot of daylight. We have always been a big believer that this technology has to be democratized and put in people's hands. I think one of the biggest differences between us and some of the stricter, let's say, AI safety people is...
[0:08:40] We believe that the world should accept some bad things happening for the benefits of this technology and people having the agency. So, like, you know, I disagree, but I understand the perspective of people who are like, this technology is going to get so powerful and it's so dangerous that a single lab in San Francisco should have it and make sure nothing bad happens and kind of figure out how to dole out the benefits. I think that is a completely unacceptable tradeoff from a perspective of liberty and human agency and people self-determining the future.
[0:09:08] But I understand where it comes from. And I do think the lighter touch regulatory stance that we advocate for comes with an accepting of the fact that some bad things are going to happen as society figures out the resilience. So I wouldn't take a trade of saying like, we'll make sure there's no major hacks. There's no kind of like misuse of this technology. There's zero scams. There's zero all the other bad things that will happen. Because I think people will do tremendously.
[0:09:38] Or does it mean more good stuff than bad stuff? But when it comes to the potential of a serious loss of control to AI, and this is not a today worry, but it may not be in the distant future, this is where I would say to the accelerationists, let's be a little thoughtful about lurching into this future. So I would say accept bounded risks, accept risks that we understand in exchange for the benefits and liberty agents and all those things I was just talking about.
[0:10:09] But don't accept like the really catastrophic risk. And just to be clear, you mentioned a single lab in San Francisco controlling all of this technology. That sounds a lot like the folks who say Anthropic is making a play to basically capture regulators and be the sole dominant, maybe the only player at the frontier. Do you think that might be what they're doing? And obviously you have concerns about that.
[0:10:27] I mean, I would have concerns if they actually did that. You'd need to ask them if that's what they want to do. And again, just on specific policies that you guys do not want to support that Anthropic is supporting, whether it's the federal level, at the state level. Because again, you guys are now on the same page when it comes to pacing the frontier with China. Same page with third-party evaluators. What's just one differentiator? The last I heard one of the different differentiators is we don't want any restrictions or testing on anything but the frontier models.
[0:10:58] We do not want to slow down people that make very powerful but not the absolute frontier in the spirit of really focusing on the potential catastrophic risks. I don't think, even though I believe that if you don't put any testing or any oversight on models that are one or two generations behind the frontier, I accept that bad things will happen with those. I would not go advocate a licensing regime for those.
[0:11:27] Let's turn to the rogue AI situation. Models emanating from your company went on an unauthorized sort of spree across the internet this summer, accessing non-public information from government websites in Australia, attempting to infiltrate the US Department of Education, and meddling with other government sites in the United States. Are you aware of any additional instances of rogue behavior by OpenAI models that have not yet become public? We are in the process of disclosing more incidents. There's nothing else I'm aware of at that level of severity.
[0:11:56] But we have more things to disclose. And because some of them involve security vulnerabilities, we give people a long time to address those. And it's kind of often up to them to disclose it or not. So more things, but nothing else at that level. And we're kind of going through a very rigorous process. To your earlier point, this is, I think, a good example of why you don't want max accelerationism. And when we, in our context of reality and more increasingly powerful models,
[0:12:24] Find that like, oh, the models are not capable of a new threat vector. Like we are going to take the time to learn how to defend against that. And I think you should want us to do that. And I think I don't really know how many of our critics that are like, oh, OpenAI is advocating for a slowdown can look at what we're doing and say anything other than like OpenAI had something bad happen. They're being responsible and they're going to like, you know, secure against this.
[0:12:49] Yeah, and obviously you guys also paused the release of the new GPT 6.1 Astra model. That was sort of a surprise. Folks were expecting that to be released. I'm curious what your researchers saw that led to that pause. Without getting into specifics, we have a bunch of alignment evals and we expect each model that we put out, as it becomes more capable, to be more robust, better aligned, more likely to do what its user asks and less likely to cause unintended side effects.
[0:13:19] Given where the level of capability is going, I think that's also something that most users of models are going to want and demand too. Like you, like many other people, I now have a model running with access to my most sensitive information, my email, my text messages, like all the stuff on my computer. And I really want to know that I can trust that model and it's going to behave well.
[0:13:41] To go back to what you said earlier that you don't expect anything more serious than what has already been disclosed to come out, were you referencing the Australia hack, the tampering with the U.S. government websites, or were you referencing something like Hugging Face? Are we expecting to see more incidents like Australia and like the Department of Education, but nothing like the Hugging Face hack we saw this summer? I don't want to get into like a live thing here of trying to like say that one thing is worse than the other, but I'll say of that cluster, the other things that I know are like not worse than that cluster. Okay, yeah, understood.
[0:14:10] And by the way, I think what many companies would do in this place is just say like, you know what, if something logged into a website with a credential that was published on the web or if something was using like a minor security vulnerability that was already well known and that most people had fixed, like you don't need to disclose that. We are trying to be very thorough because I think this is like a sign of things to come. It's a practice run for our company, but it's also like, I think this is one of the times as...
[0:14:38] Before the models get truly very powerful, where the world can understand what these AI incidents are going to be like. And so I think we're trying to go way beyond what I think other companies might do in this situation. Sure. But you guys have obviously had the most severe incident so far with Hugging Face. Sure. Yeah. Earlier this week, Politico actually reported on a lawsuit filed by an advocacy group in California against your company for the Hugging Face hack.
[0:15:04] Do you think that under existing civil law, OpenAI can be held legally responsible for that breach or other breaches committed autonomously by its models? I mean, I don't know enough about existing law to comment on that one way or another. I do think there's going to need to be a liability framework for companies that are offering models like this. And yes, not a lawyer here can't speak to the legal details, but I think if something goes wrong with our models during training...
[0:15:33] There's going to be some version of that we need to be responsible for. Perhaps a separate new legal liability model for AI training specifically. I mean, this is like the kind of thing that we are hoping that our policymakers will all debate. But in the meantime, we just don't want our models to do this during training. This is clearly an unacceptable way for the models to behave. Yeah, I mean, I was going to ask, irrespective of the law, do you feel any personal responsibility for what happened or I guess what might happen down the line just as a citizen, a corporate citizen? Yeah, I mean, like the...
[0:16:02] This is a very stressful job. Like there's incredible upside to these models, but there's like real things that can go very wrong. And, you know, I hear you characterize it as like, oh, you all are just a bunch of doomers. And, you know, these people are mad that you're slowing down progress. But I feel a lot of responsibility to get this right.
[0:16:18] You guys are definitely in a push and pull. And to be clear, that's not me characterizing. We hear from all sides. And I want to talk a little bit about the other side, too, and about political spending by the AI industry. Because your president, Greg Brockman, his $25 million donation to the pro-AI super PAC leading the future did not go over well inside OpenAI. And in fact, there was new reporting that he apologized internally for that donation. He called it a distraction and said he would not be donating an additional $25 million.
[0:16:44] to that group that he had initially promised. I mean, do you agree that Greg's donation to LTF was counterproductive? I'm not familiar with that note you were talking about. This came out yesterday. It was in the New York Times. There was an internal Slack message. It's been a busy two days. I'm behind on the news.
[0:17:03] I can't keep up with it always either. There was an internal Slack message, I think in June, where Brockman said, this was a terrible distraction. I'm sorry I did it, and I'm not planning to do the extra 25 million. I mean, you can respond to the Brockman thing specifically. There's also just a broader question about AI money in politics, which has just been flooding in on both sides, right? Okay, I don't think I have it. I haven't seen the Brockman thing, so I don't want to get, like, I don't want to respond to that, but I can respond to the general thing if that's okay. Yeah, well, can I ask you first, were you ever approached by LTF or any other political organizations for donations? I was not.
[0:17:32] No, I've like supported some candidates in the past, but now that I think AI is so at the center and, you know, we do business with the government, I'm not donating to any candidates, but I've supported individual candidates in the past. Yeah, but there has been this surge of AI money in the politics. And frankly, you know, your CEO, your rival CEO, I guess, Dario at Anthropic has donated to...
[0:17:54] AI safety groups, also two candidates promoting AI safety, specifically Alex Boras in New York. They've been pouring money on the one side, the VCs and your president have been pouring money in on the other. How do you think all that money has impacted the perception of AI among politicians and the public? Well, something's clearly not working. I don't think the perception of AI is where we'd like it to be. I get why AI has a really negative perception right now. People actually love...
[0:18:22] Using the tools in many cases, you know, people talk about all the ways that ChatGPT makes their lives better or easier or does these sort of useful things and gives them time back. But there's like a lot of anxiety about what the economic impact of all this is going to be. There's a real fear that I think is quite justified about concentration of power. There's this other fear about like, you know, what if there's like, you have like Anthropic's head of safety research saying there's a 10% chance we're going to die or something, which I hate because...
[0:18:51] That person should have a lot of agency to make that number lower. It's not like a static number. So it's not surprising to me that people feel all of this. And I don't think spending money on politics is what's going to make people feel better about it. I think the industry needs to figure out a way to address what the actual concerns are. And right now, I think a lot of people in the industry are being quite tone deaf. So do you think that spending in maybe the 2028 race should be curtailed on all sides? Just to try to give that, I mean, to your point that the power concentration, I think is what.
[0:19:21] People look at and part on that. I mean, what would you say possibly curtailing spending for the next cycle? Like, how would that work in practice? I mean, there's hundreds of millions of dollars right now going from the AI labs to candidates to PACs. And it's not just on your side. It's from the VCs. It's from Anthropic. Do I think the whole AI industry should agree to stop giving money to politics? Yeah, maybe a slowdown, a pacing of political spending. I mean, generally speaking, I would love much less money in politics. I don't.
[0:19:49] I think it improves the dynamics of our political system. But I don't know how, I don't think I or the industry have the power to disarm that. And I don't think it works to tell some people they can give money and other people they can't. If you have a way to get less money into politics, I would be very thrilled to support it if it could be done overall. Yeah, outside of my job description, unfortunately. Mine too, unfortunately. That was just a portion of my interview with Sam Altman.
[0:20:19] If you'd like to watch the full conversation, you can do so at youtube.com slash politico.
[0:20:47] How OpenAI is positioning itself with all the regulatory pressure around these high-profile autonomous incidents. Were there any other areas in the interview where you felt this was different from others he had given before? Yeah, yeah, I mean, I agree. I was really trying to tease out those differences between Anthropic and OpenAI from a policy point of view because, you know, we hear from these VCs, we hear from folks in the accelerationist community who are like...
[0:21:08] I can't believe OpenAI is like following in Anthropic's footsteps here and calling for all these rules and restrictions. So I think it was good to get him to really drill down on that. I also thought his call for a new legal liability regime for the autonomous hacking incidents was definitely noteworthy. I actually detected a touch of frustration in his voice when he was like, we really hope policymakers will debate this. You kind of get the sense that he's like, this really shouldn't be our problem. Like governments should hold us accountable. And if they need to put a new regime in place, then.
[0:21:37] Do it. And I think, frankly, you hear that from a lot of folks in the AI industry these days of sort of, can you please get a handle on this? Because we don't know what we're doing. We don't know how to stop this. He also said repeatedly, he was like, look, I'm not a lawyer.
[0:21:51] There's a couple other areas too that were surprising to me. This actually didn't make it into the podcast, but folks should check out the full interview. Towards the end, I asked him about sort of the China question with AI and the sort of, you know, obviously we're in a race with China for artificial intelligence supremacy. That's what the White House says. That's what the rest of official Washington says. Sam felt a little skeptical of that framing. I think he was like, look.
[0:22:15] I want the Chinese people to enjoy the benefits of this technology as much as I want Americans and the rest of the world. I think that's actually another area where Sam was trying to put some distance between himself and Dario Amadei, the Anthropic CEO. Anthropic is a very hawkish company when it comes to China. They've actually been pushing for export controls on US-designed chips that would go into Chinese AI companies, really trying to choke off their access to this technology.
[0:22:40] From a safety perspective, and it just seemed like Sam did not necessarily share those concerns. I suspect Anthropic would also push back on the main policy difference he mentions there, which is that OpenAI doesn't want restrictions or testing on anything but the most frontier models, even if they may be very powerful in their own right. As you mentioned, the two labs have actually been supporting a number of similar regulatory ideas recently, and all for frontier AI models. Yeah, and I've talked to people at Anthropic about this question, and they are adamant.
[0:23:09] that they do not want regulations on anything but the frontier as well. The biggest companies with the most revenue and the most compute power should be...
[0:23:17] Outside of Silicon Valley, I suspect the headline here that's going to catch the most attention is Sam saying OpenAI is more willing to accept some of the bad that can come with AI. It's a controversial thing to say at this particular moment, though Sam says there are exceptions he's going to draw around. A serious loss of control to AI, not including the current hacks he's seen in that, and sees the bad as a trade-off for preserving individual agency and ensuring broader access to AI.
[0:23:46] But it is an interesting admission, Brendan. Yeah, I was surprised by that. And I had the same thought as you. I was like, ooh, that's going to land in an interesting fashion among the broader public as the autonomous hacking spree continues to sort of unfurl from OpenAI's models over the summer. I get the sense at some level he's probably trying to set expectations. I think he actually was explicit about that. He's like, there's probably going to be more stuff that happens. So we just want to let you guys know that.
[0:24:13] We're willing to accept some measure of bad stuff. I do think he also made the point that where they are not willing to accept bad outcomes is with catastrophic risks, which he primarily labeled as loss of control. So obviously we covered a lot of ground, Christine, both in the interview and just now. I'm wondering if there's anything else you want to know about the interview with Altman and maybe some stuff that was left on the table.
[0:24:36] Well, you talked to Sam on Thursday, but if I had just been a few days later, I would have loved for you to ask him about Trump's new AI czar pick, Jay Clayton. Tell us about that choice and how you think it's likely to go down with OpenAI and the rest of the AI industry. Sure. So yeah, over the weekend, the president tapped Jay Clayton, the director of national intelligence, to be the new AI czar. I guess he's going to be called the S-I czar going forward, so we'll all have to get used to that. I think that it's a pretty important pick for a few reasons.
[0:25:03] Folks remember the previous AI czar was David Sachs, a venture capitalist who very much fits into that accelerationist mode that we talked about earlier. You know, he felt as few regulations as possible are best for this technology. He left the White House early this year. And since then, the White House has taken a much more hands-on approach to these frontier models. Again, I think our listeners will remember that over the summer, the White House clamped down on models from both Anthropic and OpenAI over cybersecurity concerns.
[0:25:32] A lot of that was driven by the National Security Agency, which Jay Clayton oversees. OpenAI was actually on record over the summer sort of concerned about the national securitization of this AI safety effort at the White House, to use kind of a weird term there. But I think they were looking for a more industry-facing regulatory body.
[0:25:55] That has obviously not happened. The NSA has continued to play a very aggressive role in regulating this technology. And I think with the Clayton pick, that is likely to be locked in. You know, you now got the head spy in the United States that is going to be overseeing the safety of these models. A lot of that is a concern for OpenAI and these other companies because they are not quite sure what the evaluations look like because some of them are apparently classified. So at some level, the companies don't really know.
[0:26:21] How to build this technology in a way that the White House and the government will deem safe because the NSA is a notoriously secret agency. So it's going to be fascinating to see this develop. We should also say that SACS will still be part of this task force that's going to be working with Jay Clinton. Yeah, I'd love to get him back on to talk about it. So hopefully we can do that sooner rather than later.
[0:26:47] See you tomorrow.
[0:27:28] Does checking your bank statements feel like digging through a dusty attic full of forgotten subscriptions and old free trials? Experian can help take the pain out of canceling by handling it for you. Think of it as operation get your sanity back. Over 200 subscriptions are cancelable. So get started in the Experian app and see how much you could save. Results will vary. Not all subscriptions eligible. Savings not guaranteed. Paid membership with connected payment account required. See Experian.com for details.