Chatbot Ads: What's Working, What Isn't
Original: Ads in Chatbots Have Arrived. Here’s What’s Working—and What’s Not
Why This Matters
Ad revenue may be the only path to sustainability for smaller AI platforms.
As OpenAI reports 100K+ ChatGPT advertisers since February and Google expands AI Mode ads to 10+ countries, ad tech startup Koah Labs offers early data on what chatbot advertising actually delivers.
Anthropic is staying ad-free—its recent IPO filing doesn't even mention advertising as a revenue source, backed by roughly $112 billion in investor capital. Everyone else is moving fast. OpenAI has surpassed 100,000 advertisers in ChatGPT since launching ads in February 2026. Google last week expanded ads in its AI Mode search chatbot to 10 more countries, claiming users find them 'highly relevant.' Advertisers, however, remain unconvinced the experience is fully there yet.
San Francisco-based Koah Labs, which places ads across 200+ AI apps including DeepAI and Liner, is building an early picture of what works. CEO Nic Baird says AI-composed ad copy generates double the clicks of fully human-written ads—likely because machine-generated text feels more natural alongside AI-generated responses. Targeting is also evolving beyond keyword matching: Koah shows spa ads after stress-related prompts by reading both the user's query and the AI's full response, pulling context from the entire conversation. That's a meaningful step past what search engines can do.
Not everything runs smoothly. Regulated advertisers like pharmaceutical companies, bound by disclosure requirements, lean toward human-controlled copy—and see weaker results. The broader risk: bland or intrusive ads could push users away from niche chatbot platforms that, unlike OpenAI or Google, can't absorb the financial hit.