The Brand Surface Has Moved
While most brand leaders are integrating AI into their apps, agents are moving into messaging. These are not the same strategy.
You might have read about Peter Steinberger in your feed last week. The AI developer who got sued over his product’s name and then sold to OpenAI.
You probably thought this was the end of the story.
It’s actually the beginning of a much bigger one.
In late 2025, Steinberger built a tool that got quietly famous among those who follow the evolution of AI.
He initially called his invention Clawdbot. It was an agent. It was clever, and it worked. And within weeks it had a devoted following of early adopters sharing it the way people used to share links to something genuinely new.
Then Anthropic, the AI company behind Claude AI sent Pete a letter. Polite, apparently, but firm. His name was too close to “Claude.” He’d have to change it.
What followed was a fast-moving branding disaster. The ten-second gap between releasing his old social media handles and claiming new ones was enough for automated bots to hijack his accounts. Scammers exploited a slip-up he made in the transition from his old social media handle to the new one and launched a fake cryptocurrency in his name. Its value quickly jumped to $16 million before collapsing, but this rocked Steinberger, who posted publicly, "I was close to crying. Everything’s fcked.”
But he persisted. And after a few transitionary names, he finally settled on OpenClaw.
Last week, Sam Altman tweeted that Steinberger would be joining his company and that OpenClaw was to “become core to our product offerings.” Zuck had apparently been calling, too, but Steinberger went with OpenAI because Altman agreed that he could operate Open Claw as an independent foundation.
When I first read this story, I thought: another talented engineer builds something smart, enters the zeitgeist, gets into some trademark trouble, but lands on his feet with a big payday.
Classic Silicon Valley arc. End of story.
But it turns out that it wasn’t the end of the story.
In fact, it’s actually the beginning of a much larger one, and it has big consequences for anyone who leads brand marketing and is trying to figure out what AI actually means for their business.
The question hiding inside the story
Here’s what almost nobody wrote about when covering Steinberger’s story.
The first name he gave his project wasn’t Clawdbot. It was WhatsApp Relay.
Not 'AI Agent Platform.' Not 'Intelligent Assistant Framework.'
WhatsApp Relay.
Most developers name their projects after what the product does. Steinberger named his after where it would live. That distinction—product versus placement—is the entire argument of this article.
Steinberger named his first prototype after the insight it was based on: people don't use WhatsApp. They live there. Three billion of them, checking it an average of 96 times per day. First thing in the morning. Last thing at night. If you want to get people to try something new, you don’t want to start by asking them to change their habits.
You don't create a new place for them to visit. You show up in the place they’re already standing.
This seems obvious when you say it aloud. But it hasn’t been obvious in practice over the last 15 years—because the entire architecture of modern digital brand strategy has been built around the opposite assumption.
A beautiful shop nobody visits
Think about what your brand has spent the last decade and a half building.
Apps. Websites. Digital experiences. Loyalty platforms. Each one is a destination—a shop you constructed and invited customers to visit. Each new technology wave became an argument for making that shop smarter and shinier. Mobile meant making your website work on a phone. Social meant building a presence where your customers were scrolling. AI, for most brands right now, means making your app more intelligent.
The underlying logic has been pretty consistent: the brand owns a digital space, and the relationship with the customer happens in that space. On that surface.
But here’s the uncomfortable reality that undermines that logic.
Most people rarely visit most brand websites. They download an app, use it once or twice, and forget it exists. The average person has around eighty apps on their phone but only uses nine of them regularly.
Nine.
Your brand’s carefully designed, AI-enhanced digital experience is competing for a slot in nine.
The brands that have genuinely earned that daily or weekly visit—banking apps, shopping platforms, streaming services—worked incredibly hard for it over many years. Customers have an established, almost involuntary reason to show up. For those brands, the app is a genuine brand surface.
For most other brands, the honest answer is that customers think of you when they need you. And when they need you, they’re not already in your app. They’re in their messages, their email, and their search bar. They’ll find you, eventually, if the need is strong enough.
But “eventually, if the need is strong enough” is not a brand relationship. It’s a transaction that happens to involve your brand.
What Steinberger built (and what OpenAI has now acquired) was a prototype for a different answer to this problem.
How OpenClaw works—in plain English
Here’s how OpenClaw works, described the way Steinberger described it himself: it’s an agent your mum can use.
That’s not modesty. It’s a design brief.
You message it the way you’d message a friend. Not through a special app. Not through a portal you’ve logged into. Through WhatsApp, or iMessage, or Telegram—whatever messaging app you already use every day. You write something like “find me a flight to Barcelona next Thursday under £300," or “what are the best-reviewed hotels near the conference center in Amsterdam”” or “remind me to call the dentist on Tuesday.”
It receives your message, figures out what you’re asking for, goes off and does the relevant things, and comes back with an answer or a completed action.
No interface to navigate. No account to create. No app to remember to open.
The community called it “Claude with hands.” That description is more precise than it sounds. Most AI tools that brands have deployed are, fundamentally, very good at answering questions. OpenClaw was built to do things. There’s a difference between a customer service chatbot that explains your return policy and an agent that processes the return while you’re still in conversation with it.
That’s what “hands” means. Not clever answers. Completed tasks.
And here’s the thing that made me sit up. Steinberger didn’t build a product. He built an answer to a question brands have been quietly failing to ask: what if the most powerful AI experience is the one that never asks customers to change their behavior at all?
Why this matters for your brand in the near future
I want to paint you a picture of what the next three to five years look like, because the pace of this change is faster than most brand strategies currently account for.
You’re the brand leader of a mid-sized hotel group. Right now, you have a website, an app, and a loyalty program, and you’ve recently integrated an AI chatbot into the booking flow. The chatbot is good. It’s doing its job. Your team is proud of it.
In three years, a meaningful share of your guests won’t interact with that chatbot.
Not because they dislike it. Because they’ll never navigate to the page it lives on.
They’ll ask their AI agent to find and book a hotel in Edinburgh for the weekend of March 14th that has a spa and allows dogs. The agent will check availability, read reviews, compare prices, and present options. The whole interaction will happen in their messaging environment—while they’re on the train, between meetings, or waiting to collect their kids.
Your hotel either appears in that result or it doesn’t. Your chatbot, excellent as it is, is entirely beside the point.
The competitive question won’t be “is our AI smarter than the competitor’s AI?” It will be something more uncomfortable: “Are we present in the environment where customers are making decisions—or have we built something impressive that requires customers to come to us?”
Apply this to retail. A customer asks their agent to find running shoes under $150, good for wide feet, and available for delivery by Thursday. The agent cross-references reviews, checks several retailers, and identifies options. Your brand either surfaces in that result or it doesn’t.
The sophistication of your on-platform AI experience is secondary. Presence—in the environment where the buying decision is already happening—is everything.
This doesn’t mean brand apps will disappear overnight. For categories where customers have built genuine daily habits around a digital property—banking, food delivery, streaming—those habits won’t dissolve quickly. But the percentage of brand interactions that happen inside brand-owned digital properties is going to shrink.
The question is whether your strategy sees it coming.
What the data was already saying
Steinberger’s work didn’t reveal a new phenomenon. He made something undeniable that the data had been saying for over a year.
Meta AI built a standalone app. They also embedded the same AI directly into WhatsApp.
The standalone app: four million daily users. The WhatsApp integration: six hundred million active AI users.
Same AI. Different surface. One hundred and fifty times the adoption. The only thing that changed was where it lived.
This isn’t a story about WhatsApp specifically. It’s a story about behavior change. Or more precisely, the avoidance of it. Customers are creatures of habit. They’ll use AI capability that fits into their existing movements. But they’ll largely ignore an AI capability that asks them to add a new stop to that journey.
Renaissance Hotels understood this when they launched their AI concierge through WhatsApp and SMS in late 2023 — not through their hotel app, which most guests don’t bother to download. It works because it exists where the guest already is. Not where the hotel wishes the guest would be.
There’s a lesson here that costs nothing to learn. Most brands are still paying to ignore it.
The question your next strategy meeting should start with
I’m not suggesting you tear up your app roadmap. For many brands, that investment is still the right one, and the AI integration you’re doing inside your own platforms will create genuine value.
But here’s a question worth adding to every AI investment conversation from here forward.
Where is our customer when they need what we offer, and is that where our AI actually lives?
If the answer is “they need us when they’re on the go, mid-task, in a messaging app” and your AI capability lives exclusively inside your website or app, you’ve found your gap. The customers who most need what you offer in those moments are currently getting their needs met by whoever shows up in the environment they’re already in. On the surface they’re already on.
That might not be yours.
Steinberger’s story read like a classic tech narrative. Smart developer. Lucky timing. Big exit. But what he demonstrated—in a chaotic, four-names-in-three-months, nearly-gave-up kind of way—was something with implications well beyond his own career.
He showed that the most useful AI is the kind that comes to you. Not just with information but with helping hands (or claws).
And once customers experience that, the kind that asks them to come to it will feel like an imposition by comparison.
The brand surface has moved. The brands that figure out how to be present on the new surface—while everyone else is still optimizing the old one—will look like geniuses in retrospect. The way the early mobile-first brands did. The way the first brands that built genuine social communities did.
The window for being early on this is open. But it probably won’t stay open long.