Is AI Killing Search and SEO? (cacm.acm.org)

OpenAI’s ChatGPT hit 700 million weekly active users in August 2025 — a fourfold increase year-over-year — and generative AI is now driving hundreds of millions of daily web visits (≈250M/day, ~84% from ChatGPT). That shift is already eroding traditional search dynamics: Google’s market share fell below 90%, and studies (Pew) show AI summaries in search cut downstream clicks (8% click-through vs. 15% without summaries). For publishers and ad-driven sites this threatens traffic and monetization as search results increasingly return AI-generated answers or “AI Overviews” that satisfy queries without sending users on to source pages. But SEO is not dead so much as evolving. Experts say the new battleground is “generative engine optimization” (GEO): structuring concise, machine-readable content and distributing it across channels so LLMs and retrieval systems can find and use it. Technical implications include the continued importance of traditional SEO for retrieval-augmented generation (RAG) and hybrid systems that supplement LLMs with real-time web searches, while new challenges center on appearing not only in retrieval results but in the selection step LLMs use to craft answers. Practical limits remain — direct placement inside fully synthesized responses is hard, slow to influence because of training cycles, and risks being treated as data poisoning — so publishers must adapt tactics, measurement, and expectations for an AI-first search era.