When Anthropic ran its Super Bowl commercial earlier this year, it made a bold, heartwarming promise: Claude, its AI assistant, would never be cluttered with advertisements. That promise still stands today, and it’s worth pausing to appreciate just how rare that is in the current tech landscape. The company’s recent IPO filing didn’t even mention advertising as a potential revenue stream for Claude, which tells you how firmly they’re committed to the ad-free path. It’s easy to maintain that kind of high ground when you’re sitting on roughly $112 billion in investor cash, of course. But for the rest of the chatbot industry, the math is a lot more uncomfortable. Generating smart, conversational AI responses is expensive—far more expensive than serving up a traditional search results page. Every prompt requires computing power, energy, and infrastructure, and those costs add up quickly. For companies that don’t have Anthropic’s luxury of near-bottomless funding, the question isn’t whether to make money from their chatbots, but how. And increasingly, the answer looks like the same thing that’s funded the internet for decades: advertising. That doesn’t mean every AI chatbot is about to turn into a billboard, but it does mean the era of pure, ad-free AI assistance is probably going to be a premium experience rather than the default.
OpenAI has already embraced that reality with open arms. Since ads debuted in ChatGPT back in February, the company has said that over 100,000 advertisers have signed up to place their messages inside the chatbot’s responses. That’s a staggering number, especially for a product that many people still think of as a clean, subscription-based tool. Google, meanwhile, is moving just as aggressively. Last week, it expanded advertising in its search chatbot, AI Mode, to ten more countries beyond the United States, and the company was quick to tell advertisers that users are finding these AI-generated ads “highly relevant.” Whether that’s entirely true is another matter. Advertisers themselves are far from completely satisfied with how the experience works right now. They’re still figuring out how to measure performance, how to avoid looking tone-deaf next to an AI answer, and how to make ads feel helpful rather than intrusive. But the important thing is that they expect it to improve. They’re not running away; they’re sticking around, iterating, and investing. That’s a strong signal that AI-powered ads are not just a temporary experiment. They are becoming a permanent part of the digital advertising ecosystem, and the only real question is how quickly and how gracefully that evolution happens.
For niche chatbot platforms, this trend isn’t just interesting—it’s existential. Smaller AI companies can’t rely on the same flood of venture capital that the big labs enjoy, and they certainly don’t have the massive advertising revenues of a Google or Meta to fall back on. They need to find a way to cover costs, and ads are one of the few realistic options available. That’s where companies like Koah Labs come in. Koah is a San Francisco startup that’s essentially building the plumbing for advertising inside AI experiences. It markets itself as a way to bring ads to more than 200 different apps, including DeepAI, which focuses on creative queries, and Liner, which emphasizes citation-rich answers for research-minded users. The brands that work with Koah range from telehealth giant Hims to flower delivery service UrbanStems, so you can see the variety of advertisers interested in getting in front of AI users. Koah has attracted only about a tenth of the advertisers that OpenAI has, but its CEO, Nic Baird, argues that his platform is already giving him a clear view of what’s actually working in this new frontier—and what’s still fuzzy. He’s watching chatbots replace traditional search engines in real time, and he’s learning how that shift will shape not just marketing budgets, but the very experience you have when you ask an AI for help.
One of the most interesting findings Baird shared involves who—or what—is actually writing the ads. According to him, ads that are composed partly by AI, rather than fully by a human, generate double the click-through rates. That’s a surprising stat, and even Baird acknowledges it’s difficult to say exactly why it happens. His best guess is fairly intuitive: AI-generated marketing copy simply feels more natural when it appears next to the output of an AI model. There’s a kind of tonal consistency that makes the ad blend in with the content instead of feeling like a jarring interruption. What’s perhaps more important is that these AI-crafted ads aren’t turning into hallucination factories. Because Koah grounds the ads in product details and imagery supplied directly by the advertiser, users don’t seem to be running into fabricated ingredients, fake discounts, or made-up product claims. But that doesn’t mean everything is smooth sailing. Some advertisers, particularly drugmakers, face strict government requirements about what they must say and what they cannot say in their marketing. Those companies can’t just let an AI model write whatever feels natural. They need human oversight to ensure every claim is compliant, and according to Baird, that extra control tends to generate weaker results. It’s a fascinating trade-off between safety and engagement, and right now it’s one of the biggest practical challenges facing AI advertising.
The other major advantage of AI advertising is creative targeting. With a traditional Google search, if someone types “help with stress,” the algorithm can do its best to show a relevant therapist ad, but that’s about it. Chatbots can do much more. Baird says that in Koah’s network, spa ads have performed exceptionally well when they appear after users submit prompts related to stress. That might sound simple, but the magic lies in how the ad gets selected. Koah doesn’t just look at the user’s prompt in isolation. It also looks at the AI platform’s response, which carries context from the entire conversation up to that point. So if you’ve been chatting for a while about your job, your sleep habits, and your upcoming vacation, the AI has a much richer understanding of what you might actually need than any search engine ever could. That depth of context is something search ads simply can’t match. And because chatbot prompts tend to be at least twice as long as typical Google queries, there’s just more information to work with. The potential to serve ads that feel genuinely helpful rather than randomly targeted is enormous. Instead of seeing a generic ad for a stress-relief app, you might see an ad for a weekend spa package, a weighted blanket, or a meditation retreat—something that actually matches the nuance of your situation. That doesn’t mean it will always be perfect, but it’s a level of personalization that feels genuinely new.
All of this points to a future where the line between content and advertising becomes even more blurred—and where the experience you have with AI is increasingly shaped by who is paying for it. On one side, you have Anthropic, promising a sanctuary of ad-free intelligence for those willing to pay a premium. On the other, you have a rapidly expanding ecosystem of free and low-cost chatbots that will rely on ads to keep the lights on. For users, this could go one of two ways. It could mean a web full of irritating, intrusive, and repetitive ads that make AI assistants feel less like helpful companions and more like telemarketers. Or, if companies like Koah get it right, it could mean ads that actually add value—suggestions that arrive at the right moment, in the right context, with just the right tone. The companies that succeed will be the ones that find a middle ground, creating ads that are engaging without feeling manipulative, and that protect the user experience without starving the platform of revenue. That’s a difficult balance, especially when humans are increasingly taken out of the creative loop. But if the early numbers are any indication, the machinery is already learning. We’re at the very beginning of a strange marriage between artificial intelligence and advertising, and both industries are stumbling forward together. The winners won’t necessarily be the ones who get to ads first, but the ones who figure out how to make an ad feel unmistakably human.
