[ROUNDTABLE] Anthropic (ads) vs. OpenAI (ads in services)

Ahead of Super Bowl LIX tomorrow, there appears to be a battle on the AI field before the battle on the football field.

What’s happened so far:

January 16, 2026 — OpenAI announces ads coming to ChatGPT

• The company announces it will begin testing ads within ChatGPT “in the coming weeks”

• Ads will appear at the bottom of ChatGPT’s responses for free and Go tier users

• Plus, Pro, Business and Enterprise subscriptions will remain ad-free

• Company asks select advertisers for $200,000 minimum commitment for beta testing

February 5, 2026 — Anthropic releases Super Bowl ad campaign

• Anthropic unveils four ads titled “Betrayal,” “Deception,” “Treachery,” and “Violation“

• Ads parody intrusive advertising in AI conversations, showing chatbots that form relationships with users before hawking products

• All ads end with: “Ads are coming to AI. But not to Claude”

• Two of the four ads scheduled to air during Super Bowl LIX tomorrow

February 5, 2026 (same day) — OpenAI CEO Sam Altman responds

• Altman calls Anthropic’s ads “clearly dishonest” and accuses them of being “deceptive”

• Defends OpenAI’s model: “Anthropic serves an expensive product to rich people. We need to bring AI to billions of people who can’t pay for subscriptions”

• Accuses Anthropic of being “authoritarian” and wanting to “control what people do with AI”

• States OpenAI’s own Super Bowl ad is “about builders, and how anyone can now build anything”

February 6, 2026 — OpenAI Ad testing begins

• For some brands, ad testing starts as early as this date

What now?

To better understand what matters here and what doesn’t, I called up some of my favorite minds in tech to give their thoughts. There’s so many different ways to look at this evolving dynamic (not just Anthropic ‘owning’ OpenAI), and I hope these perspectives in conversation around this rapidly evolving topic give you a better sense of what’s important.

Welcome to Cities Decoded: Roundtable edition.

The Daily TEA creator (AI and crypto news, AI memes)

ChatGPT is primarily a consumer-facing product, not an enterprise tool. OpenAI is currently struggling with economics. Its revenue can’t even cover the daily inference costs, let alone operations or other expenses. Many insiders suggest OpenAI is burning cash, and the situation is worsening as leadership shifts focus and resources toward consumer-facing ChatGPT rather than foundational research.

Internally, researchers and scientists are frustrated. They claim that compute resources are being prioritized for ChatGPT instead of R&D. OpenAI, once known as a research-driven AI lab, is becoming more of a consumer technology company—one that’s struggling to balance its books while competitors continue to innovate rapidly.

Altman once said that ads would be OpenAI’s “last fort.” That now appears to be the case. Unlike traditional software companies, where costs are minimal and profit margins are huge, AI businesses operate more like manufacturing: every product has a real cost. In AI, inference costs are unavoidable—there’s no free lunch.

Anthropic, by contrast, has taken a more sustainable path. Its main customers are enterprises, especially those using Claude for coding, and this segment pays well.

This ChatGPT ad saga is a reminder that revenue matters. Survival must come before progress. If advertising is truly OpenAI’s “last fort,” it suggests the company is focused on survival rather than its original mission of achieving AGI. Or perhaps, in a twist, OpenAI has already achieved AGI—Ads Generated Income.

Read the latest from the Daily TEA, ‘Benchmarks, Coders, Reviews, and Agents,’ here.

Jonathan Bensimon — field notes (NY tech events)

The ads are a pretty baller move by Anthropic. Not only is Claude now taking over ChatGPT/OpenAI in this arms race in many different ways, But now they’re pretty much rubbing it in that they don’t need to do any sort of advertising on their platform too, which is pretty massive. I think Claude now not only has the product factor, but also the ‘cool’ factor out of many of these platforms now.

It’s pretty awesome to see. Claude was pretty irrelevant over the last year, and so for them to pick up steam, especially like this, it’s definitely a big right hook. I’m interested to see how the other players are going to be coming back, because it’s a pretty severe blow. I’m taking out my popcorn, watching and seeing what happens next. They’re definitely in the pack.

Read the latest from jon, ‘📍AI video demos, quotes with CEO of Smooth Media and founder of CreatorNYC,’ here.

elif nisa polat — humanbeyondtech (community and AI)

Anthropic is making a statement about not going into images and video. Instead, it is focusing on becoming the workflow agent for consumers/enterprises. I respect that.

This ad feels like the good old Coca Cola vs Pepsi ads: Clever, thought provoking and builds trust among creatives who have been amongst the most vocal about what we are missing with AI slop.

People aren’t opposed to the innovation of AI. Innovation is exciting and cool for most. What they’re repulsed is the lack of authenticity in AI advertising. When the advertising misses the nuances and contexts of local communities. AI giving generic answers wont touch peoples hearts and make them buy a product.

I’m all in for AI image + video generation, as long as the artist, the creative director, the marketing agency has put enough thought and soul into their output. Check this beautiful ai marketing collection from just another Turkish woman with taste 🤜🤛🧿

Read the latest from elif, ‘wait—”hacking” isn’t always a bad thing?,’ here.

Austin Nelleson — COMPUTE/COMPETE (geopolitical AI)

These ads are the heating up of the monetization race. OpenAI is known for being the most overstretched—no big tech background and having spent the most on data centers, etc, so it now has to find miracle ways to make money.

Anthropic has found success differentiating itself (known as THE coding tool), while OpenAI continues to be the company associated with chatbots. While that is great for the ChatGPT brand, it will be hard to generate revenue from primarily consumer tools. Ads are a good start.

Read the latest from austin, ‘COMPUTE/COMPETE #1,’ here.

Hillary Zeng — Scaffold (Gen Z stories, critical thinking and AI)

What stood out to me about the Anthropic ads wasn’t the strange turn of events at the end of the video (where the supposed embodiment of ChatGPT shifts from being a helper to an advertiser). It was everything else that came before—the unnatural conversation between the man in the blue tank top and the man in the grey t-shirt.

It is so glaringly obvious that the man in the blue tank top is ChatGPT in this interpersonal interaction. The exchange feels unnatural, and the simple reason is that the blue tank top man is far too positive. No normal human is so happy, so willing to be helpful. It is this quality that distinguishes “ChatGPT” (alienates it). But isn’t it at least a bit strange for Claude to play into the narrative that optimistic/helpful is bad? And also, for Claude—the supposedly more ethical chatbot—to play into the narrative that ChatGPT is like a human being?

Read the latest from hill, ‘Testing the Prompt and Brainstorming the Interface,’ here.

Magan Chin — FIG (AI-powered career platform)

I see why OpenAI is running ads. In a way, the company has to because of how expensive it is to train their models and they’re trying to run education programs for entire countries and service entire sovereign governments and be accessible to everyone. But also, ads are obviously not ideal. So I think that’s why we need both Claude and ChatGPT to exist. I think it makes sense to have two different models!

See magan’s app, FIG, in action here.

Myself (Cities Decoded)

Say what you will, but the ads are a very, very, clever play by Anthropic. I say that not because of the vastly Anthropic-positive reaction online, but because of how quickly it provoked a thorough and defensive response from Altman.

Even though the path of AI innovation rests on the work of tens of thousands of people around the world (accompanied by their AI agents), it is still inherently a path paved by big personalities.

When there is a new breakthrough, people want to know what Musk thinks. What Thiel thinks. And in this ad-for-ad battle, there is no one I want to hear more from than Altman and Amodei.

To be clear, I don’t think the biggest figures in tech are always making good progress for humanity. But I find it fascinating how they think of this AI moment, especially when technological progress (Claude Code, Codex, etc) meets the business realities of powering such tools (ads).

I would love to see a response from Amodei, even if it is just a victory lap. But regardless of what ads either company runs tomorrow during the big game, there’s already a clear winner in my book.

Long live Anthropic…for now.

Read/watch my latest, ‘My answers to the Interact fellowship application. And why you should answer them too,’ here.

Claude ads in question

BETRAYAL:

DECEPTION:

TREACHERY:

VIOLATION: