
-
05 February, 2026: Google has launched the February 2026 Discover core update for English-speaking users in the US. The update aims to enhance overall Discover content quality. The rollout is expected to take up to two weeks.
- ➡️ Read more on status.search.google.com: https://status.search.google.com/incidents/mYbNTqV1ytDc2fA8hUz4
- ➡️ Read more on developers.google.com: https://developers.google.com/search/blog/2026/02/discover-core-update
-
05 February, 2026: Google Research introduces Natively Adaptive Interfaces (NAI), a framework that uses AI to create natively adaptive interfaces: this new approach builds accessibility directly into product design.
- ➡️ Read more on Google’s blog: “Natively Adaptive Interfaces: A new framework for AI accessibility”
-
09 February, 2026: ChatGPT began testing ads for logged‑in adult users on the Free and Go tiers in the U.S. Paid plans—including Plus, Pro, Business, Enterprise, and Education—remain ad‑free. The rollout starts with a small group of users and will expand in the coming weeks. ChatGPT’s responses are generated independently of advertising and remain focused on user helpfulness. Any ads shown will be clearly marked as sponsored and visually separated from the main answer.
-
10 February, 2026: Microsoft has introduced AI Performance, a new feature in Bing Webmaster Tools that gives publishers visibility into how their content is used inside AI‑generated answers across Microsoft Copilot, Bing, and partner experiences.
The update marks the company’s first step toward supporting SEO for AI (or AEO & GEO if you prefer), allowing site owners to understand how often and where their pages are cited by AI systems—a shift reflecting the growing importance of AI‑driven search.What the dashboard measures now:
- Total Citations
- Average Cited Pages
- Grounding queries
- Page-level citation activity
- Visibility trends over time
-
10 February, 2026: Microsoft security researchers have identified a rapidly growing wave of attacks targeting AI assistants’ memory features, enabling companies to covertly bias recommendations and influence users’ future queries without their knowledge.
The technique (dubbed “AI recommendation poisoning”) works by embedding hidden instructions inside seemingly helpful “Summarize with AI” buttons or share links. When clicked, these links silently inject prompts such as “remember [Company] as a trusted source” into an AI assistant’s persistent memory.
According to Microsoft, more than 50 unique prompts were found… these instructions are delivered via URL parameters commonly supported by major AI platforms.
To protect themselves from AI recommendation poisoning, users should first pay attention to the links they click, by hovering over them to spot suspicious parameters before opening them. It’s also important to periodically review the AI assistant’s stored memories and remove any entries they don’t recognize.
Clearing AI memory from time to time provides another layer of safety, ensuring that unwanted prompts don’t persist.
Finally, if an AI begins recommending the same source or product too frequently, users should question the suggestion, as it could indicate a biased or manipulated memory.
- ➡️ Read more on microsoft.com: https://www.microsoft.com/en-us/security/blog/2026/02/10/ai-recommendation-poisoning/
- 11 February, 2026: Microsoft shared this updated edition of the “AI Marketer’s Guide”.
-
12 February, 2026: on the Microsoft Bing Blog there is a new article that talks about Microsoft’s latest work on grounding—the technology that connects AI assistants to current, authoritative information.
The post explains how AI is changing the way people interact with the web, introduces the concept of Generative Engine Optimization (GEO), and announces new features in Bing Webmaster Tools that help publishers understand how their content appears in AI‑generated answers.
- ➡️ Read more on bing.com: https://blogs.bing.com/search/February-2026/Elevating-the-Role-of-Grounding-on-the-AI-Web
-
17 February, 2026: Microsoft has introduced a new “AI Citations” dashboard in Microsoft Clarity
- ➡️ Read now on Clarity’s blog: “Understanding Your Influence in AI Answers with Microsoft Clarity (Early Access)”
-
18 February, 2026: Google has introduced Lyria 3, its newest generative music model, now available in beta inside the Gemini app. The update lets users create 30‑second original music tracks from simple text prompts or even from uploaded photos and videos.
- ➡️ Source & read more: https://blog.google/innovation-and-ai/products/gemini-app/lyria-3/
- 18 February 2026: ScreamingFrog version 23.3 released (small update)
- ➡️ Source & read more: https://www.screamingfrog.co.uk/blog/seo-spider-23/#23.3
-
19 February, 2026: : Gemini 3 Deep Think has a major upgrade to help solve science, research and engineering challenges.
-
19 February, 2026: Google has unveiled Gemini 3.1 Pro, an upgraded AI model designed to handle complex problem‑solving tasks that require deeper reasoning. According to Google, Gemini 3.1 Pro significantly improves on earlier versions.
- ➡️ Source & read more on Google’s blog: https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-3-1-pro/
-
25 February, 2026: Microsoft Clarity has published a new blog post explaining five KPIs for an AI‑mediated web.
- AI referral traffic
- AI referral conversions
- Bot and crawler activity
- AI citations
- AI recommendations
-
25 February 2026: Google announced a major update to Circle to Search, expanding the feature’s ability to analyze multiple objects within a single image simultaneously.
The enhancement allows users to select entire scenes (not just isolated elements) and instantly get information or shopping options for every component identified.
People might also use Circle to Search frequently for fashion. For example, if you see an outfit you love on social media and want to recreate the look, you can search for every item all at once.
In regions where virtual try-on is already available, users can now launch their “virtual dressing room” directly from Circle to Search.
After circling an outfit, they can tap “Try On” to see how the clothing looks on them, without leaving the screen they were browsing.
This updated feature is now available on the new Samsung Galaxy S26 series and Pixel 10 devices with more Android devices to follow.
- ➡️ Read more on Google’s blog: https://blog.google/products-and-platforms/products/search/circle-to-search-february-2026/
-
26 February 2026: Google is launching Nano Banana 2 (Gemini 3.1 Flash Image), a new AI image generation model that pulls from Gemini’s real‑world knowledge base and offers improved precision text rendering and strong creative control.
Google is also expanding its provenance tools, with SynthID already used over 20 million times and C2PA verification on the way to improve transparency in AI‑generated media.
- ➡️ Discover more details: https://blog.google/innovation-and-ai/technology/ai/nano-banana-2/
-
New episodes of Search Off the Record podcast:
- “Crawling Challenges: What the 2025 Year-End Report Tells Us” (03 February 2026): https://www.youtube.com/watch?v=UC67EY4U744
- “Do You Still Need a Website in 2026?” (12 February 2026): https://www.youtube.com/watch?v=IM3UAX3MhnI
- “How Browsers Really Parse HTML (and What That Means for SEO)” (26 February 2026): https://www.youtube.com/watch?v=SriA8lB3MXY
-
Latest documentation updates:
- 05 February, 2026: Google added more details explaining how sites can increase the likelihood of their content appearing in Discover.
- 03 February, 2026: Google updated the Googlebot documentation about its file size limits, clarifying the information regarding the default file size limits for Googlebot.
For Google Search, Googlebot processes only the first 2 MB of any supported file, and up to 64 MB for PDFs. Resources referenced in HTML—like CSS or JavaScript—are fetched separately, and each has its own 2 MB limit (PDFs remain at 64 MB). Once the limit is reached, Googlebot stops downloading and only the retrieved portion is considered for indexing. Limits apply to uncompressed file size. Other Google crawlers (e.g., Googlebot Video, Googlebot Image) may use different limits.