Five focused guides for publishers and brand teams. See what changed, check the source and decide what to measure next.
Start with one buyer question. Record whether Google shows an AI Overview and whether it cites a page that answers the question. Then run the same question in ChatGPT Search and compare the source URLs.
Test Claude Fable separately as a tool for reviewing evidence and long documents. Its release changes which model a team can test for that work. It does not tell you which pages Google or ChatGPT will cite.
The November rack
Hover to turn a cover · drag the rail · click to openSwipe the rail · tap a cover to open it
Short answer
Each story is answer-first, links to its original sources, and comes in two formats: a full article and a 60-second story.
Claude Fable 5.2 is Anthropic’s model for demanding reasoning and long-running technical work. Anthropic recommends starting with Opus 5.5 for most coding and knowledge tasks, then testing Fable on the hardest cases.
In November 2026, OpenAI's consumer product lead said ChatGPT's search-style queries are about 20% of Google's volume. OpenAI is also testing a Search Console-like analytics tool for verified businesses inside a Business Hub. The figure is OpenAI's own claim and the tool has not launched.
A September Ahrefs study found AI Overviews on 82.9% of tracked branded U.S. desktop search results on September 29. November’s Google AI search briefing covers that finding alongside changes to source links, advertising and access.
Perplexity Teammates, announced in early November 2026, are AI agents that keep working between prompts and return finished output for review. Each Teammate gets its own desktop and browser and can be reached in Slack. It's a web research preview for Perplexity Max ($200/month).
Use five checks: record AI Overview citations, measure ChatGPT Search separately, validate analytics availability, make important facts readable, and distinguish ads from earned links.
Key numbers
The figures behind the stories, with the source for each.
| Number | What it measures | Story | Source |
|---|---|---|---|
| 1M | Token context window | Claude Fable 5.2 is available | Anthropic |
| Read-only | No money can be moved | Claude Personal Finance is live (in beta) | TechCrunch |
| ~20% | OpenAI estimate of relative query volume | OpenAI estimates a one-fifth ratio | Big Technology |
| Web | Only web-backed queries count | What counts as a search-like query? | Big Technology |
| 83% | Tracked branded U.S. desktop results on Sep 29 | Coverage varies by query, market and device | Ahrefs |
| 4 | ad slots per AI Mode answer | Up to four ads in AI Mode | PPC Land |
| 24/7 | Ongoing work | Agents that work between prompts | Tech Insider |
| 1:1 | Agent : browser | Each Teammate gets its own desktop and browser | Tech Insider |
| 83% | Tracked branded U.S. desktop results on Sep 29 | Measure AI Overviews on your buyer queries | Ahrefs |
| ~20% | OpenAI estimate of relative query volume | Treat ChatGPT as a search channel | Big Technology |
Platforms covered
Previous edition: AI News, July 2026. For the long-term picture, see AI search and GEO statistics for 2026.
FAQ
One question per story, answered directly.
Anthropic recommends Opus 5.5 for most coding and knowledge tasks. Try Fable 5.2 for demanding, open-ended work where a controlled test shows enough improvement to justify the extra cost.
No. By OpenAI's own figure (November 2026), ChatGPT search-style queries are about 20% of Google's volume. The comparison's Google baseline was not disclosed.
Coverage depends on what is measured. Ahrefs found AI Overviews on about 83% of tracked branded U.S. desktop results on September 29, 2026.
AI agents from Perplexity, announced in November 2026, that keep working between prompts, each with its own desktop and browser, and that can be reached through Slack.
GEO is the practice of making a brand visible, cited and recommended in AI-generated answers from engines like ChatGPT, Gemini, Claude, Perplexity and Google AI Overviews.
Arrow AI
The free audit shows where ChatGPT, Claude, Gemini, Perplexity and AI Overviews already mention you, and where competitors are cited instead.
Explore the topic: AI search & AI news
