December is shaping up as the first full month of ‘Gemini 3-era’ search, with Google now auto‑routing tougher AI Mode queries to its frontier model and quietly reshaping which entities get surfaced in answers. AI Overviews are testing bigger safety disclaimers, new event layouts, and deeper travel/booking workflows – all of which keep users inside Google’s generative layer longer. At the same time, ChatGPT and Perplexity have leaned hard into visual answers and agentic shopping, turning AI search into an end‑to‑end discovery + checkout funnel. Expect December to be about one thing: can your brand be selected by AI systems, not just ranked in blue links?
(All developments below are based on changes and public reports from Nov 21–30, 2025.)
After launching Gemini 3 on November 18, Google flipped the next switch on November 26: automatic model routing in AI Mode now sends “your toughest questions” straight to Gemini 3 Pro. Previously, only Pro/Ultra users who picked the “Thinking” option could force Gemini 3. Now, Google decides per query. For SEOs, this means complex, multi‑step queries are more likely to trigger:
Deeper query fan‑out (more background searches),
Richer generative layouts (tables, tools, simulations),
And potentially new pages being pulled in that never ranked top‑10 before.
Treat AI Mode as a parallel discovery system: optimize entities, schema, and depth so your content is “Gemini‑ready,” not just “rank‑ready.”
The inevitable happened: ads are now showing inside AI Mode answers. Greg Sterling and Brodie Clark spotted sponsored cards at the bottom of Gemini‑powered AI responses, with Google confirming these are part of ongoing ad tests first announced in May. The current pattern:
Organic link cards appear around / above the AI answer,
Sponsored tiles sit at the bottom of the AI block, labeled as ads.
Right now, organic visibility isn’t being displaced at the top of AI Mode – but this is clearly the monetization blueprint. Expect Google to iterate on positioning, formats, and targeting. SEOs need to watch how often their commercial SERPs now show AI Mode plus ads, not just classic ads + ten blue links.
Google is testing a much larger “Double‑check important information” notice at the bottom of some AI Overviews, telling users to verify critical info and offering a “Learn more” link. This seems to trigger more often on YMYL queries (finance, health, etc.). It reframes AI Overviews as a starting point, not final authority.
At the same time, AI Overviews are experimenting with a new direct event display, replacing the traditional event module with AI‑driven event cards – currently buggy, but live for some users. For local/event SEOs, that means structured event data and strong entity connections matter twice: to rank in Events and to be selected for AI Overviews’ event layouts.
Late November saw another burst of ranking chaos, with a “Thanksgiving update” reported around November 24–26 on top of the broader “Movember” volatility earlier in the month. Multiple roundups describe:
No official core update, but big swings across many verticals,
Ongoing crackdowns on site reputation abuse and low‑quality third‑party content,
AI Overviews appearing in roughly 30% of queries in some datasets.
Combined with Q3 data showing AI Overviews can reduce organic CTR by ~61% and paid CTR by ~68% on affected queries, we’re clearly in a world where ranking volatility + zero‑click behavior are now normal conditions, not anomalies.
While rankings wobble, Search Console finally got more “AI‑era” instrumentation:
Branded Queries Filter (announced Nov 20; rolling out now) automatically splits branded vs non‑branded traffic in Performance reports using Google’s own AI classification.
Custom Annotations in performance charts let you log site changes, algorithm updates, migrations, and even your AI experiments inside GSC instead of in spreadsheets.
Query Groups in Search Console Insights use AI to cluster similar queries by topic and show trending up/down query families.
Action item: set a December baseline. Annotate all AI‑related tests (new schema, FAQ pruning, IA changes) and start reporting non‑brand vs brand SEO impact as separate storylines.
OpenAI’s November releases turned ChatGPT into an even stronger “answer + search” surface:
Nov 21 – More images in answers: ChatGPT now injects inline web images for people, places, and products, with clickable sources, when visuals add clarity.
Nov 24 – Shopping research: a new interactive flow where ChatGPT deeply researches products, asks clarifying questions, and returns personalized buyer guides with links to buy. This is live on web and mobile for Free, Go, Plus, and Pro users with generous holiday limits.
This turns ChatGPT into a visual, commerce‑aware answer engine, not just a chatbox. For SEOs, it increases the importance of structured product data, high‑quality images, and brand coverage across the open web that ChatGPT crawls and cites.
Perplexity spent late November turning itself into an AI‑first shopping and browsing environment:
It launched an AI personal shopper that lets users search conversationally for products (“best winter jacket if I live in San Francisco and take a ferry to work”) and get personalized cards with specs and reviews.
Purchases can be completed via PayPal Instant Buy inside Perplexity, with a holiday promo offering 50% back (up to $50) for early US users who pay with PayPal.
The Comet AI browser is now on Android, bringing its agentic, task‑based browsing and multi‑tab AI research workflows to mobile.
SEO implication: product discovery is drifting into closed AI funnels where citations + commerce integrations matter more than ten blue links. Treat Perplexity as a channel to monitor, not a curiosity.
Barry Schwartz
Barry spent the last week chronicling how Google’s AI search stack is shifting in real time: Gemini 3’s auto‑routing in AI Mode, new “double‑check important information” flags in AI Overviews, bugs in event displays, OpenAI’s ramp‑up of crawling IPs, and the unconfirmed Thanksgiving ranking update. His take is clear: SEOs are now dealing with “always‑on” micro‑updates across rankings, SERP layouts, and AI behavior – with very little official communication.
Brodie Clark
Brodie was one of the first to trigger and document ads inside AI Mode, sharing screenshots that show sponsored tiles at the bottom of Gemini’s answer block. He contrasted the new layout with earlier AI Overview ad tests and highlighted that, for now, organic link cards still sit above the ads. For practitioners, his message is blunt: AI Mode isn’t just a UX experiment anymore; it’s entering the paid search strategy deck and will demand new SERP feature tracking.
Sharon McClintic
In Lumar’s November 2025 SEO & AI Search roundup, Sharon framed the month as a pivot point from “rank tracking” to “AI visibility tracking.” She connects Gemini 3’s launch, Comet’s Android expansion, new GSC branded filters, and fresh CTR studies into one narrative: AI layers are changing how and where clicks happen, not just which position you hold. Her advice leans entity‑first: map your brand’s presence across AI Overviews, AI Mode, ChatGPT, Claude, and Perplexity instead of obsessing over one SERP.
Duane Martinez
Duane’s November 21 analysis of Google Search Ranking Updates argues that 2025 marks the real shift from classic SEO to Generative Engine Optimization (GEO). He documents the “Movember” volatility, estimates AI Overviews now show on ~30% of queries, and highlights enforcement against site‑reputation abuse and affiliate‑heavy pages. His playbook: lean into authority, structured data, and genuinely helpful content that’s easy for AI systems to cite, not just easy for algorithms to rank.
Ryan Siddle
Ryan, from Merj, surfaced a quieter but important story: OpenAI is scaling up its crawling infrastructure for the holiday season, adding many new IP blocks for its bots. His takeaway for technical SEOs: update WAF rules and bot‑detection logic so you can distinguish OpenAI from generic scrapers, and watch how often AI crawlers hit key templates. It’s a reminder that AI visibility starts in your logs — not just in Google Search Console.
Vijay Chauhan
Event SEO specialist Vijay Chauhan spotted Google’s test of direct event displays inside AI Overviews, where AI now renders event entities in the overview itself instead of handing them off to the classic Events module. He notes that the layout is currently buggy, but significance is clear: AI Overviews are becoming a self‑contained canvas for local discovery. For event and venue sites, Vijay’s experiments suggest doubling down on clean event schema, unique images, and strong entity links to city and venue nodes.
Valentin Pletzer
Valentin was among the first to catch the new “Double‑check important information” notice in AI Overviews and share proof. His screenshots highlight how much more prominent this warning is than the older, subtle disclaimers. The subtext for SEOs: Google is visibly distancing itself from AI mistakes, especially on YMYL topics. That likely increases the bar for which sources are eligible for AI citations – expect more weight on expertise signals, consensus across sources, and clear disclaimers in your own content.
Expect automatic Gemini 3 routing to expand beyond AI Mode and deeper into AI Overviews, especially for complex or multi‑step queries. This should increase the importance of entity‑rich pages that answer entire tasks, not single keywords.
With ads already live in AI Mode tests and Google openly experimenting with placements, December will likely bring broader exposure and more formats. Watch for new ad cards in AI Overviews and start modeling how much your commercial traffic shifts into AI+ads blends.
The branded queries filter, custom annotations, and Query Groups together give you a new reporting stack. Use December to: separate brand vs non‑brand SEO, annotate every AI‑related change, and measure topic‑level volatility instead of relying on single keywords.
ChatGPT’s shopping research and Perplexity’s PayPal‑powered personal shopper show that AI platforms want to own the full shopping funnel. At the same time, OpenAI is scaling up crawling, meaning more of your content will feed non‑Google ecosystems. In December, prioritize: robust structured data, canonical clarity, and AI‑specific crawl controls (robots/llms files, WAF rules).
👉 Follow Manikandan N on LinkedIn for weekly insights on AI-powered SEO, entity optimisation, and how India’s search landscape is evolving with generative AI.
Stay ahead of Google updates, AI search experiments, and SEO automation workflows with insights from global top SEO minds. No noise - just what matters, delivered every week.