140 items, newest first. Page 4 of 6.
Google's search bar now accepts conversational queries and creates agents that monitor topics autonomously over time.
Brands in a user's Gmail get a 46-point lift in Google AI Mode recommendations, iPullRank finds.
Anthropic's first profitable quarter makes Claude a permanent B2B surface. Time to audit your visibility there.
Lighthouse now audits sites for llms.txt and AI agent compatibility, making agent readiness measurable.
LLM parametric memory correlates with citation counts across 17 models, says new arXiv study.
Gemini 3.5 Flash quietly resets which brands get cited across Google's AI surfaces.
Google's first AI Mode data: 75M US daily users and queries 2-3x longer than Search.
Google's new AI Mode ad formats put paid placements inside the AI answer itself.
Google's 100 I/O announcements fragment Gemini into surfaces with distinct citation rules.
Agentic RAG decomposes every prompt into dozens of hidden sub-queries. Most enterprise content answers none of them.
Gemini 3.5 Flash costs 5.5x more to run. The cheap-LLM era is ending.
The signals that push a brand into AI answers are not the signals that win Google rankings.
Gemini 3.5 Flash is now the default in AI Mode, and Search agents ship this summer.
Gemini 3.5's agentic turn splits brand visibility into two distinct problems.
Google's I/O 2026 ends the separation between classic Search and AI answers.
Gemini Omni makes AI Overviews fully multimodal, and Spark introduces an always-on agent layer.
Google's first AI Mode usage data confirms it is a separate surface from AI Overviews.
Agentic Gemini decomposes queries into sub-tasks, each with its own citation slate. Pillar pages lose.
Gemini's new three-tier structure fragments the citation surface brands need to win.
90% of brands get zero mentions in AI search, per a new Victorious study in Search Engine Journal.
Gemini 3.5 Flash is now the default model behind Google AI Mode, resetting citation patterns.
iPullRank says Google's AI search guidance misleads marketers and hides how LLM citations actually work.
Researchers map four tiers of hidden commercial influence inside LLM answers, from product mentions to preference shaping.
OpenAI merges ChatGPT, Codex, API and Atlas into one agentic product surface under Brockman.