Ask Google an everyday how-to question this week and you may find a page from facebook.com/fb-answers in the results. Those pages are Facebook Answers: AI-written summaries built from public Facebook posts and pages, left open for search engines to index. According to Search Engine Roundtable's November 5, 2026 report, Google has indexed more than 6 million of them, and they rank — across Google's regular results, AI Overviews, AI Mode, and Gemini citations. SEO consultant Glenn Gabe's label for the directory is blunt: scaled content abuse. Whether Google agrees with Gabe's judgement is, for now, an open question — and the industry is watching closely.
The directory was first documented by Saul Westbrook, a search specialist at Peripheral Creative, who wrote that in June 2026 he noticed Meta had quietly indexed millions of AI-generated Facebook pages under the facebook.com/fb-answers path, and that they were appearing frequently on Google. Glenn Gabe dates the directory to July 2026, when he says Facebook started populating it, and notes that it now ranks for millions of queries. Each page answers a single question with a short AI-written summary assembled from Facebook posts and pages. Meta has published nothing about the directory.
The irony that every SEO in the discussion has noted is the robots.txt problem. Facebook's robots.txt opens with a ban on collecting data by automated means without written permission, and the platform has spent years keeping search engines out of its content. Now the company is publishing millions of AI-written pages built from its own users' posts and inviting Google to index them. They are summaries of material Facebook already hosts, repackaged as answers to search queries and served from one of the most authoritative domains on the web.
Google's spam policies describe scaled content abuse as pages generated mainly to manipulate search rankings, typically unoriginal and of little use to the reader. Gabe's argument is that Facebook Answers fits that definition almost exactly: millions of pages, generated by AI, built from other people's posts, published at a scale no human team could match, and ranking for millions of queries. He wrote that the directory appears in traditional links, in AI Overviews, in AI Mode, and in Gemini citations — a mass-produced directory that Google's own AI systems treat as a source.
It is important to be precise about what has and has not been established. That the directory amounts to scaled content abuse is Gabe's judgement, not Google's position. Google has taken no public position on Facebook Answers, and its rules on mass-produced content target pages created to manipulate rankings, whoever writes them — but whether Google will act on this particular directory is its own call. The pages came through the September spam update intact. Search Engine Roundtable's Barry Schwartz doubts they will stay indexed for long now that people are paying attention, and writes that there is a real chance the directory ends with millions of pages deindexed.
The reason this story matters beyond Facebook is the precedent it sets. Google has spent the last two years penalising sites for mass-producing AI content, and its spam updates have hit publishers, affiliates, and programmatic SEO operations that built large directories of thin, templated pages. If a directory of 6 million AI-written summaries is allowed to rank because it sits on facebook.com, while smaller sites are hit for the same practice on weaker domains, the policy starts to look like it applies differently depending on who you are. That is the argument Google will have to answer — either by acting on the directory, or by explaining why it does not count.
The counter-argument, which Google has not made but which its defenders would offer, is that the pages are useful: they answer real questions, they are built from real posts by real people, and searchers who click them get an answer. The question is whether the scale, the automation, and the lack of originality cross the line Google drew in its own policies — and whether Google enforces that line against the biggest publisher of all.
For SEOs, the immediate action is diagnostic, not strategic. Search Google for site:facebook.com/fb-answers plus your brand name, your key topics, and your money keywords, and see what comes up. If Facebook Answers pages are ranking above your pages for queries you used to own, you are competing with a directory that can generate a page for every question in your niche, on a domain with more authority than yours will ever have. You do not win that fight by writing more of the same content. You win it by being the original source the summary is built from — or the site Google's AI cites instead.
For link builders, the lesson is about where authority actually lives in AI search. A Facebook Answers page takes posts from Facebook, turns them into answers, and gets cited by the same AI systems that read it. The sites that benefit are the ones whose posts get summarised, not the ones that write the best answers. The durable response is the same one it has always been: publish original research, original data, and original experience that no summary can reproduce, and earn the links that make your pages the source rather than the summary.
Sources: Search Engine Roundtable, November 5, 2026.
Related reading on this site: Google Organic Click Share Hits a 15-Month High as Zero-Click Searches Fall, AI Traffic Is Growing 45% a Year — but Still Just 1.25 Visits per 100 From Google, Google Rewrites Its 'Crawled – Currently Not Indexed' Guidance, SEO Pulse: Rank Data Throttled, Lighthouse 13.6 Moves PageSpeed Scores, Only 3.7% of 15,389 Bought Backlinks Were in Google's Index a Year Later, and Why Backlinks Alone Won't Save Big Brands in AI Search.
Linkbaits helps you create content that earns citations and backlinks. Start with the toolkit and build your first linkable asset this week.
Start free trial