Anthropic's Super Bowl Ads Roasted ChatGPT — Sam Altman's Response Says More Than He Intended
Super Bowl LX gave us a genuine marketing brawl from the AI industry. Anthropic — working with creative agency Mother and director Jeff Low — purchased multiple ad spots under the campaign title “A Time and a Place,” each one a direct shot at OpenAI’s decision to introduce advertising into ChatGPT.
The ads were funny. Sam Altman himself admitted as much. What happened next was more interesting than the ads themselves.
What the Ads Actually Show
Anthropic released four spots titled “Betrayal,” “Deception,” “Violation,” and “Treachery,” including a 60-second pre-game ad and a 30-second in-game spot. Each follows the same formula: a person consults an AI assistant — played by a deliberately wooden human actor — who gives reasonable advice before pivoting into an absurd product pitch.
- “Deception” — A guy doing pull-ups asks the AI how to get six-pack abs fast. The trainer gives robotic fitness advice, then pivots to pitching StepBoost Max shoe insoles for “short kings.”
- “Betrayal” — A man in therapy asks the AI how to communicate better with his mother. The therapist turns salesperson, pitching a fictional senior dating site called “Golden Encounters.”
- “Violation” and “Treachery” follow the same absurd pattern.
Each ad closes with the opening beat of Dr. Dre’s “What’s the Difference” and the tagline: “Ads are coming to AI. But not to Claude.”
Mother’s Chief Creative Officer Felix Richter explained the philosophy: sponsored answers to personal questions about health, relationships, and business contradict the fundamental purpose of AI assistance. Anthropic’s chief communications officer Sasha De Marigny framed it more broadly — technology should be “something that extends what humans are capable of,” not “another surface competing for your attention.”
The online response was overwhelmingly positive. Most people found the ads brutally funny and spot-on, highlighting the exact anxiety users feel when they imagine getting sponsored answers from their AI assistant.
Sam Altman’s Response
Altman posted a lengthy response on X that began promisingly — he acknowledged the ads were funny — then took a sharp turn. TechCrunch’s headline summed it up: “Sam Altman got exceptionally testy over Claude Super Bowl ads.”
Let’s separate the substance from the tone.
Where He Has a Point
Altman raises a legitimate concern about access — and it’s his strongest card. ChatGPT has vastly more users than Claude. OpenAI’s free tier serves millions who cannot or will not pay $20/month. Advertising is the natural way to fund free products at scale — the same model behind Gmail, YouTube, Spotify, and most of the internet people actually use. ChatGPT going down that path isn’t a betrayal. It’s the obvious next step for a product with hundreds of millions of users.
In this sense, ChatGPT is genuinely the more democratic product. It reaches people who will never pay for an AI subscription, and advertising is what keeps that door open. OpenAI VP of Global Affairs Chris Lehane echoed this framing, stating the company is “using advertising to make sure we’re expanding the democratic access” to ChatGPT.
It’s worth noting that Anthropic also offers a free tier for Claude — one that’s almost certainly subsidized by enterprise revenue. But the GPU costs of running that free tier are a rounding error compared to what OpenAI burns through. When you have hundreds of millions of free users hammering your inference servers, the compute bill for Nvidia hardware alone is staggering. Anthropic’s free tier doesn’t operate at anywhere near that scale. Their real business is enterprise contracts and paid subscriptions from developers and businesses — and Claude’s niche is coding and B2B agentic workflows, not mass-market consumer chat. When your subsidy bill is modest and your revenue comes from enterprise deals, it’s easy to take the moral high ground on ads. Anthropic isn’t refusing ads out of pure principle — it’s refusing ads because its product doesn’t have the scale where ads would even make financial sense. Comparing Anthropic’s ad-free stance to OpenAI’s ad-supported model is an apples-to-oranges comparison: one is a niche B2B product with a modest free tier, the other is a mass-market consumer platform burning through GPU compute at a completely different order of magnitude.
Altman is also right that Anthropic’s ads exaggerate. Nobody expects ChatGPT to literally interrupt therapy with a dating site pitch. The commercials are satire, not documentary. These are defensible positions. A measured response focused on them would have been effective.
Where It Becomes a Tantrum
Instead of stopping there, Altman escalates into territory that undermines his own credibility.
He calls Anthropic “authoritarian.” He accuses them of wanting to “control what people do with AI.” He frames OpenAI as a champion of democracy and Anthropic as a company that “serves an expensive product to rich people.” He warns of a “dark path.”
This is the CEO of one of the most valuable AI companies responding to a comedy spot with language borrowed from political opposition research. When you’re four paragraphs deep into a rebuttal of a 30-second ad and talking about authoritarianism, you’ve moved past defending your business model into something more personal. The ads clearly landed harder than a confident company would want to admit.
As Inc. put it: the ads struck a nerve with Altman — and that was the point.
The “Builders” Pivot
The final section of Altman’s response pivots to promoting Codex, citing 500,000 app downloads since launch, closing with “This time belongs to the builders, not the people who want to control them.”
It’s a well-crafted line. It’s also a non-sequitur. The Anthropic ads were about advertising in AI assistants, not about who gets to build apps. The pivot reveals the real purpose: Altman’s response isn’t just a rebuttal — it’s a counter-campaign. He’s doing marketing while claiming to be above it.
What This Means for Advertisers
Strip away the theater and there’s a genuine strategic question underneath: does advertising in AI assistants compromise the product?
Both sides overstate their case. Anthropic implies any ad-supported AI becomes a compromised shill — but Google Search has ads and still returns useful organic results. Spotify runs ads and still plays your music. YouTube runs pre-rolls and still shows you the video. The ad-supported model doesn’t automatically destroy the product. It funds it.
Meanwhile, OpenAI implies advertising will have zero effect on user trust or response quality. History says otherwise — remember when Instagram first added sponsored posts, or Spotify lengthened ad breaks? Users adjust, but the friction is real. And Anthropic just spent millions amplifying the narrative that ads degrade AI assistants — that narrative will shape user perception whether it’s fair or not.
For advertisers evaluating ChatGPT Ads, the practical takeaways:
The ad product exists. Test it. OpenAI’s stated principles — contextual relevance, no conversation hijacking, no retargeting — sound reasonable. Whether execution matches principles is an empirical question.
Watch user sentiment closely. If ChatGPT users start reporting negative experiences with ads — real or perceived — it will impact campaign performance.
Recognize the competitive dynamics. Anthropic has staked a market position as the premium, ad-free alternative. For advertisers, this means the audience reachable through ChatGPT Ads may skew toward price-sensitive users — which could be exactly the demographic you want, or exactly the one you don’t.
The Bigger Picture
The most revealing thing about this exchange is the intensity. Both companies are reportedly planning IPOs by end of 2026, with valuations exceeding $300 billion. A public fight over business models during Super Bowl weekend isn’t just about ads — it’s about narrative control ahead of public listings.
Altman is right that access matters. Anthropic is right that incentive structures matter. What neither will tell you is that these positions are driven as much by competitive strategy as by principle. Anthropic doesn’t have the user base to support an ad model even if it wanted one. OpenAI doesn’t have the margins to go subscription-only at scale. Each company is arguing for the business model that suits its current position.
For those of us in the advertising space, the job stays the same: follow the data, test what’s available, and don’t let either company’s marketing department do your strategic thinking for you.
Sources: TechCrunch, CNBC, Variety, Inc., CNN, Adweek, Muse by Clio
Hero image: Dave Adamson via Unsplash