OpenAI Bought a Talk Show. The Org Chart Tells You Everything.

Last week, OpenAI paid somewhere in the "low hundreds of millions"⁸ for TBPN, a daily three-hour tech talk show with 58,000 YouTube subscribers and 11 employees. The show generated about $5 million in ad revenue in 2025¹. On the math alone, this is absurd. OpenAI, which just closed a $122 billion funding round at an $852 billion valuation⁶, is paying nine figures for a startup whose annual revenue wouldn't cover one month of the company's cloud computing bill.

So what did OpenAI actually buy?

The answer is in the org chart. TBPN will report to Chris Lehane⁴, OpenAI's chief global affairs officer. Lehane is a veteran political operative, the man who coined the phrase "vast right-wing conspiracy" during his Clinton White House days and who later architected Fairshake⁴, the crypto industry super PAC that spent hundreds of millions targeting anti-crypto congressional candidates in 2024. The show was not placed inside a content division or a product group. It was placed inside OpenAI's Strategy organization, alongside the company's lobbying and policy apparatus. As one analysis put it, "placing an editorially 'independent' media acquisition inside the strategy office of your chief political operative is not a subtle signal"⁹.

OpenAI has been emphatic about editorial independence. Fidji Simo, the company's CEO of Applications, said TBPN will "continue to run their programming, choose their guests, and make their own editorial decisions"⁵. Sam Altman posted on X that he doesn't "expect them to go any easier on us." I take them at their word that no one at OpenAI will pick up a red phone and tell the hosts what to say.

But that's not how media influence works, and OpenAI knows it.

The show's DNA was never adversarial. TBPN's hosts, John Coogan and Jordi Hays, are former startup founders, not journalists. They built a show that NPR described³ as "a high-energy mix of friendly interviews with tech titans, industry gossip, and celebrations of funding rounds." Elizabeth Spiers, the media strategist who co-founded Gawker, observed³ that the hosts "generally believe that most of what's happening in Silicon Valley is a good thing for society." The hosts themselves described the acquisition as a move from "commentary to real impact"¹ — language that reveals the shift from observer to participant. You don't need to censor a show that already agrees with you. You just need to own it.

As part of the deal, TBPN is shutting down its advertising business entirely⁷. The hosts will also contribute to OpenAI's broader communications and marketing efforts⁷ beyond the show. So to be clear about the structure: an editorially "independent" media property, stripped of its own revenue, staffed by people who also do marketing for OpenAI, reporting to the company's chief political operative. I think you can see the tension.

Why now? The timing is the most revealing part. OpenAI is approaching an IPO. It faces intensifying competition from Google, Anthropic, and xAI¹¹. And the public has turned sharply skeptical. An NBC News poll from March 2026¹⁰ found that 57% of registered voters believe the risks of AI outweigh the benefits, and AI's net favorability rating is lower than ICE, Donald Trump, or Kamala Harris. Only Iran and the Democratic Party poll worse. As tech scholar Sara M. Watson told NPR³: "Popular opinion has shifted to say, 'We're actually quite skeptical of your claims.'"

OpenAI is not just fighting for market share. It is fighting for permission to exist at the scale it aspires to. And the TBPN acquisition is part of a broader pattern that makes the stakes clear.

Consider what OpenAI has been doing on the regulatory front. In its formal submission to the White House Office of Science and Technology Policy¹⁴, OpenAI requested federal preemption of state AI laws in exchange for voluntary data sharing — asking, in essence, for "relief from the 781 and counting proposed AI-related bills already introduced this year in US states"¹². Chris Lehane has been lobbying the Trump administration⁴ for policies that would prevent states from regulating AI and ease environmental restrictions on data centers. A super PAC backed by OpenAI president Greg Brockman¹³ is spending heavily to elect candidates who oppose AI regulation, following what MIT Technology Review calls "the crypto industry's playbook."

Now add a media property to that toolkit. OpenAI is simultaneously (1) a regulated entity subject to pending AI legislation, (2) a party actively lobbying for specific regulatory frameworks that would benefit it over competitors, and (3) the owner of a media property whose primary subject is AI and the tech industry. These three roles are in direct structural tension, regardless of anyone's good intentions.

The strongest counterargument is that this is nothing new. Jeff Bezos owns the Washington Post. Salesforce's Marc Benioff owns Time. General Electric once owned NBC. Media ownership by corporations with political interests is as old as media itself, and the information environment is pluralistic enough that one talk show cannot monopolize the conversation.

I think this argument is correct at the level of principle and wrong at the level of specifics. The Washington Post covers general news. NBC covered everything. TBPN covers exactly the same subject that OpenAI is lobbying about, to exactly the audience of founders, VCs, and executives who shape the industry's direction. The overlap between the media property's editorial scope and the owner's regulatory interests is near-total, and that makes it structurally different from a defense contractor owning a general-interest broadcast network. Wharton M&A professor Paul Nary put it plainly¹¹: "We'll give you editorial control, but you'll still be involved in our company. So is there a conflict of interest there?"

There is a further irony that may undo the whole strategy. TBPN's value to OpenAI depends on competitors continuing to send their executives on the show. Zuckerberg, Nadella, and Benioff have all appeared. But as multiple analysts have noted⁷, executives from Google, Anthropic, and Meta may now hesitate to participate on a platform owned by their chief rival. If they stop coming, TBPN becomes an in-house podcast dressed up as journalism — precisely the fate that killed Andreessen Horowitz's Future.com, which "sputtered out in less than a year and a half"¹⁵ after launching with similar ambitions in 2021.

I think the TBPN acquisition is strategically rational for OpenAI and structurally corrosive for the information ecosystem. Both things are true simultaneously, and the second matters more. We are in a moment where the definitions of "safe AI" and "responsible regulation" are being written, and those definitions will shape which companies survive and which products reach the market. Having any single commercially interested party own a chunk of the infrastructure through which those definitions are publicly debated is a problem that contractual editorial independence cannot solve. The incentives run too deep, and they operate without anyone needing to make a phone call.

Here is what I am watching. First, whether Google, Anthropic, and Meta continue sending executives onto TBPN over the next six months — if they don't, the show's value collapses and we will learn quickly whether OpenAI repurposes it as a corporate content channel. Second, whether TBPN covers OpenAI's upcoming IPO with the same skepticism an independent outlet would, particularly around the company's $8.5 billion in annualized losses¹⁷ and the growing difficulty selling OpenAI shares on secondary markets¹⁶. Third, whether other AI companies follow suit — Plaid already bought This Week in Fintech², and if this becomes a pattern, we will end up in a world where every major AI company owns the media that covers it. That is a world with a lot of content and very little journalism, and the people who lose are the ones making decisions about a technology that 57% of Americans already think is too risky.