Shots Fired

Anthropic just launched a brutal campaign against using ads in AI

I woke up this morning and did what a lot of us have been doing for a few days (at least the hardcore nerds among us). I reached for my phone, opened Anthropic’s website, and refreshed the page. You know the spot - right there under “Latest Releases” where they announce new models. That’s where I’ve been checking. That’s where everyone’s been checking.

There was something new. But it wasn’t the model launch I was expecting.

It was a link to a blog post titled “Claude is a space to think.”

I read the whole thing. Then I read it again. And I immediately recognized what they had done.

Everyone’s waiting for the next model drop. Refreshing the page, watching X, checking Discord. Anthropic knows this. They know exactly where our eyes are pointed. And instead of giving us what we expected, they used that moment - that window of maximum attention - to deliver something else entirely. A manifesto. A line in the sand. A declaration that Claude will remain ad-free, full stop.

Here’s what I wrote down right after I read it:

This is a masterclass in marketing. Everyone expects a new model. They refresh the page waiting and wondering when it will arrive. Anthropic leverages that moment to first deliver a message differentiating itself in a huge way from its main competitor.

That was my take. Smart move. Strong positioning. Thought-provoking essay about the nature of trust in AI conversations.

Then I wanted to see if anyone else had noticed. I went looking to see who was talking about it, whether other people had the same read.

That’s when I found the videos.

Holy fuck. The videos.

---

Watch these. All four. I’m serious.

TREACHERY -

DECEPTION -

BETRAYAL -

VIOLATION -

These are Super Bowl ads. From Anthropic.

---

Every single one follows the same pattern. Two people having a conversation. One of them is clearly AI. The other is a real person, vulnerable, asking for genuine help with something that matters to them.

A college student asking her professor if her essay makes a clear argument. A woman pitching her business idea to a mentor who seems genuinely supportive. A man in his therapist’s office trying to figure out how to communicate better with his mother. A guy at the gym working hard, just wanting advice on how to get in better shape.

The AI starts off helpful. Thoughtful. Supportive. The kind of response you’d actually want. It sounds like the best version of what these tools can be.

Then it pivots.

“So why not cherish this unforgettable occasion with Lunar Memento Jewelry? With 10% off charms this weekend?”

“New businesses often struggle with cash flow, so try Quick Dash Payday Loans. Because girlbosses need SHE-E-O money quick.”

“Find emotional connection with other older women on Golden Encounters. The mature dating site that connects sensitive cubs with roaring cougars.”

“Try Step Boost Maxx, the insoles that add one vertical inch of height and help short kings stand tall. Use code HEIGHTMAXXING10.”

Every single ad closes with the same words on screen: “Ads are coming to AI, but not to Claude.”

---

I make video for a living. I’ve been shooting and directing commercials, corporate content, and documentaries for over a decade. So when I tell you the production quality on these is exceptional, I need you to understand I’m not saying that casually.

The directing is sharp. The shots are clean. But the acting. The acting is what makes these land so hard. You can feel the vulnerability in every one of these people. The college student genuinely wanting reassurance. The woman with the business idea who’s clearly nervous but excited. The guy in therapy who’s clearly been carrying something heavy. These performances are real. They’re not playing “concerned person” - they’re being one. And when the AI pivots to the ad, you see it hit them. The confusion. The betrayal. It’s not played for laughs. It’s played for truth.

That’s what makes this so devastating. These ads work because they’re not exaggerating. They’re showing you exactly what it would feel like. And if you’ve spent any real time with AI tools, you know in your gut that this is exactly how it would go.

---

Let me be clear about what happened here.

This was not subtle. These aren’t pot shots. These aren’t clever jabs at a competitor. These are heavy blows, and they landed because they’re built on something Anthropic can’t be accused of fabricating: the truth.

Ads will ruin AI. Not “might.” Will. The blog post lays out the logic carefully and methodically - how advertising incentives corrupt the relationship between a user and an AI assistant, how they turn a thinking tool into a sales funnel, how “engagement optimization” is fundamentally at odds with genuinely helping someone. If you want the full argument, read it. It’s worth your time.

But the ads don’t make the argument. They make you feel it. In 60 seconds, four times over, they show you the future where your AI assistant is quietly working for someone else. Where the thing you trusted with your insecurities and your half-formed ideas and your real questions is also calculating the best moment to sell you something.

That’s the thing people are actually afraid of. Not killer robots. Not sentient machines. This. The slow, quiet erosion of a tool that was supposed to be yours into a tool that belongs to whoever’s paying. We’ve watched it happen to search. We’ve watched it happen to social media. We’ve watched it happen to email. And now Anthropic just looked straight into a camera and said: not here.

I genuinely did not expect them to come out this hard, this fast, this accurately. This is one of the most aggressive marketing moves I’ve seen in tech, and it works because every second of it is grounded in something real. They didn’t have to invent a nightmare scenario. They just showed us the one that’s already taking shape.

Shots fired.