Research & sources
Official documentation, original papers and clearly labeled secondary evidence — with unresolved questions preserved as research gaps.
Source registry
See the agents, set the terms: new visibility and control for personal agents
Cloudflare
Original source ↗Archived evidence ↗
Completeness and Cost of Retrieval Structures for Multi-Tool Requests in LLM Agents
arXiv
Original source ↗Archived evidence ↗
Your content, your choice: clearer preferences for training and grounding
Microsoft Bing
Original source ↗Archived evidence ↗
A Sequential Decision Framework for Evaluating LLM-Guided Search
arXiv
Original source ↗Archived evidence ↗
ATLAS-Bench: An Agentic Tool-Use Benchmark for Local Area Search
arXiv
Original source ↗Archived evidence ↗
Does the Citation Matter? Quality, Grounding, and Physician-Rubric Accuracy of Google's AI Overviews for Medical Information
arXiv
Original source ↗Archived evidence ↗
Who Owns Digital News, and Who Gets Cited: Measuring Media Concentration Across 77 Countries and Seven AI Systems
arXiv
Original source ↗Archived evidence ↗
Bringing shopping into the future: The Fall '26 Edition for developers
Shopify
Original source ↗Archived evidence ↗
Contextual Embeddings Beyond the Gold Chunk
Perplexity and turbopuffer research authors
Original source ↗Archived evidence ↗
Predicting What Answer Engines Cite: Open Decision Models, Rerankers and a Distilled 4B Student on Ten Engines under Controlled Retrieval
arXiv
Original source ↗Archived evidence ↗
Who wins when AI does the shopping? New evidence on agentic commerce
McKinsey & Company
Original source ↗Archived evidence ↗
PailitaoMMSearch2.0 Technical Report: Scaling Generative Retrieval to Unified End-to-End Multimodal Search
arXiv; Taobao Pailitao industrial research
Original source ↗Archived evidence ↗
ACARE: Agent-Context-Aware Reranker for Iterative Search
Original research authors / arXiv
Original source ↗Archived evidence ↗
SearchHorizonBench: A Benchmark for Search Value in AI Agents
arXiv
Original source ↗Archived evidence ↗
ChatGPT Ads expands to 26 new countries as over 100,000 advertisers have used the platform
OpenAI
Original source ↗Archived evidence ↗
How Commerce.com and Cloudflare are getting storefronts ready for AI agents
Cloudflare
Original source ↗Archived evidence ↗
How we built ChatGPT’s novel ad stack for user trust and relevance
OpenAI
Original source ↗Archived evidence ↗
Walmart Makes It Possible for Agents to Shop on Behalf of Customers, While Keeping Customers in Control
Walmart Inc.
Original source ↗Archived evidence ↗
GLOBE: Evolving Structured Agentic Memory for E-commerce Product Optimization in Generative Search
arXiv; LG Electronics AI Lab
Original source ↗Archived evidence ↗
Location, Location, Generation: Where You Ask Matters for Generative Engine Optimization
arXiv
Original source ↗Archived evidence ↗
Tool Recommendation for Multi-Step Agents: Finding Every Tool a Task Needs
arXiv / Arcade.dev
Original source ↗Archived evidence ↗
Cloudflare and OpenAI: Making AI discovery better for everyone
Cloudflare and OpenAI
Original source ↗Archived evidence ↗
Do Agents Need a New Kind of Paper? SPRA: Structured Paper Representation for Agents
arXiv; Shanghai Artificial Intelligence Laboratory / Shanghai Jiao Tong University
Original source ↗Archived evidence ↗
Learning When to Think: Utility-Guided Adaptive Reasoning for Efficient Text Retrieval
arXiv; accepted to Findings of EMNLP 2026
Original source ↗Archived evidence ↗
Retrieved but Not Read: Understanding Why Search-Augmented LLMs Fail to Use Available Evidence
arXiv
Original source ↗Archived evidence ↗
Introducing Topics Portfolio (Beta) in AI Performance
Microsoft Bing
Original source ↗Archived evidence ↗
Plan or Retrieve First? A Scale Study of Tool Discovery Architectures
arXiv / Apple-affiliated authors; accepted at EMNLP 2026 Industry Track
Original source ↗Archived evidence ↗
Glasseek: A Transparent Recipe for Training and Trustworthy Evaluation of DeepSearch Agents
arXiv
Original source ↗Archived evidence ↗
Understanding ChatGPT Citations: Evidence from Production Web Search Across Model Generations
arXiv (original preprint)
Original source ↗Archived evidence ↗
What Drives AI Citations? Content, Retrieval and Model Knowledge in Generative Engine Optimization (GEO)
Mahe Chen and Nick Koudas / arXiv
Original source ↗Archived evidence ↗
From Public Posts to AI-Search Citations: Measuring the Fragility of AI Search
Qi Liu et al. / arXiv
Original source ↗Archived evidence ↗
From Delivery to Stateful Exploration: Rethinking the Index for Agentic Search
arXiv
Original source ↗Archived evidence ↗
Beyond Semantic Similarity: Performance and Costs of Agentic Retrieval for Complex Tasks
arXiv
Original source ↗Archived evidence ↗
Programmatic Search Agents: Extending Agentic Search Beyond Query Reformulation
arXiv
Original source ↗Archived evidence ↗
T-Search: An Open Agentic Retriever and Playground for Hard Multi-Step Search
arXiv / T-Tech
Original source ↗Archived evidence ↗
Source Preference in the Wild: How LLM Agents Favor Items by Source, and How to Reduce It
arXiv
Original source ↗Archived evidence ↗
Answer Bubbles: Information Exposure in AI-Mediated Search
Association for Computational Linguistics
Original source ↗Archived evidence ↗
Authority Bias in Conversational Search Engines for Academic Paper Recommendation.
Association for Computational Linguistics, EMNLP 2026
Original source ↗Archived evidence ↗
EvidenT: Building Trustworthy Enterprise Assistants through Evidence Groundedness and Traceability
Association for Computational Linguistics
Original source ↗Archived evidence ↗
MCite-RL: Towards Reliable Multimodal RAG via Citation-enhanced Agentic Reinforcement Learning
Association for Computational Linguistics
Original source ↗Archived evidence ↗
ReCite: Agentic Reasoning for Faithful Citation
Association for Computational Linguistics, Findings of EMNLP 2026
Original source ↗Archived evidence ↗
The Bottleneck Is Retrieval, Not Reading: Diagnosing Evidence-Based Generation with RAG for Long Context Question Answering
Association for Computational Linguistics
Original source ↗Archived evidence ↗
When Deep Research Agents Stagnate: Enhancing Reasoning with Retrieval-Aware Agent Control
Association for Computational Linguistics
Original source ↗Archived evidence ↗
Who is the Agent to Blame? Localizing Faithfulness and Citation Mistakes in Agentic Deep Research
Association for Computational Linguistics, EMNLP 2026
Original source ↗Archived evidence ↗
Breaking News Out of the Filter Bubble: Generative AI Search Diversifies Collective Attention and Raises Shared Information Consumption
Original research authors / arXiv
Original source ↗Archived evidence ↗
Learning to Route in Visual Space via Multi-Step Embedding Retrieval
arXiv / UT Austin and Google DeepMind-affiliated research
Original source ↗Archived evidence ↗
AI-based semantic search improves e-commerce product search more consistently when augmenting keyword matching than when replacing it
Scientific Reports (Nature Portfolio)
Original source ↗Archived evidence ↗
Inspire: Benchmarking Scientific Literature Search for Open Research Problems
arXiv
Original source ↗Archived evidence ↗
Announcing web multimodal Search performance reporting in Search Console
Google Search Central
Original source ↗Archived evidence ↗
Photon: Building a retrieval and ranking engine from scratch
Perplexity Engineering
Original source ↗Archived evidence ↗
Shopping by algorithm: How agentic AI deploys human heuristics as a surrogate consumer
arXiv
Original source ↗Archived evidence ↗
Beyond Relevance: Structured Semantic Supervision for Product Search with LLM-Augmented Annotations
arXiv; University of Surrey and collaborators
Original source ↗Archived evidence ↗
Scoring With the Engine: Retrieval Exposure, Cross-Engine Divergence, and the Limits of Engine-Agnostic GEO Scores
arXiv / Benjamin Tannenbaum
Original source ↗Archived evidence ↗
Lighthouse v13.5.0 adds Agent Resource Discovery audit
GoogleChrome Lighthouse
Original source ↗Archived evidence ↗
About Universal Commerce Protocol (UCP) integration hub in Merchant Center
Google for Developers
Original source ↗Characterizing Web Search by Conversational LLM Agents: From Search Decisions and Strategies to Results and Responses
arXiv
Original source ↗Archived evidence ↗
Bypassing inference bottlenecks: Accelerating complex AI search with Retrieve-for-Train
Google Research
Original source ↗Google Tests Paying Publishers For AI Answers Via Search Console
Search Engine Journal
Original source ↗Have it both ways: stay discoverable in search while disallowing AI training
Cloudflare
Original source ↗Archived evidence ↗
Introducing Gemini 3.8 Live and 3.8 Live Extended Thinking
Google / Google DeepMind
Original source ↗Archived evidence ↗
CITECHOICE: A Causal Audit of How Document Presentation Redistributes Citation Credit in Agentic Search
arXiv
Original source ↗Archived evidence ↗
CobbleDB: Rebuilding AI Search Storage for Lower Latency and Cost
Perplexity
Original source ↗Archived evidence ↗
CobbleDB: Rebuilding AI Search Storage for Lower Latency and Cost
Perplexity Research
Original source ↗AI Search Visibility Report for August 2026: LinkedIn enters the top three
Meltwater
Original source ↗Generative Marketing Mix Modeling: A Causal Inference Framework Linking GEO and GEM to Business Impact
arXiv
Original source ↗Archived evidence ↗
Help your readers find your site through preferred sources in Google Search
Google Search Central
Original source ↗Archived evidence ↗
Scoring Without the Engine: Validating a Deterministic, Manipulation-Resistant Content Score for Generative Engines, End to End
arXiv
Original source ↗Archived evidence ↗
Counter-GEO-Bench: Evaluating Defenses Against Information-Distorting Generative Engine Optimization
arXiv
Original source ↗Archived evidence ↗
When Optimization Becomes Manipulation: Defending Generative Search against Malicious Generative Engine Optimization
arXiv
Original source ↗Archived evidence ↗
Web Bot Auth HTTP Message Signatures Protocol
IETF Web Bot Auth Working Group
Original source ↗Archived evidence ↗
CHASE: How Content Ecosystems Are Reshaped When Ranking Is the Only Target
arXiv
Original source ↗Archived evidence ↗
Agent2UCB: Agentic System for Generative Engine Optimization
arXiv / original research authors
Original source ↗Archived evidence ↗
BotBase for Operators: A clearer path to joining Cloudflare's directory of bots and agents
Cloudflare
Original source ↗Beyond the Vacuum: Combinatorial Strategy Selection for Competitor-Aware Generative Engine Optimization
arXiv
Original source ↗Archived evidence ↗
AI in Search Reduces Publisher Referrals Without Improving User Experience: Experimental Evidence
arXiv
Original source ↗GEO-Flag: Detecting and Measuring GEO-Optimized Web Content
arXiv
Original source ↗Archived evidence ↗
WANDR: A Benchmark for Wide and Deep Research
arXiv / Perplexity Research
Original source ↗Archived evidence ↗
Mechanism Design for Generative Engines: From Exploitation toward Win-Win Outcomes
arXiv
Original source ↗Archived evidence ↗
Measuring Visibility in the AI Era
Interactive Advertising Bureau (IAB)
Original source ↗Archived evidence ↗
Search, Inspect, Fetch: Exploiting Boolean Retrieval for Deep-Research Agents
arXiv
Original source ↗Archived evidence ↗
EMBL AI Librarian: Life-Sciences Knowledge Layer for AI Agents
arXiv; EMBL-associated authors
Original source ↗Archived evidence ↗
Platform properties roll out globally, plus a new social and video performance guide
Google Search Central
Original source ↗Archived evidence ↗
What Do Chinese-Language Generative Search Engines Cite and Surface? A Large-Scale Empirical Study
arXiv
Original source ↗Archived evidence ↗
SearchOS-V1: Towards Robust Open-Domain Information-Seeking Agent Collaboration
Renmin University of China / Ant Group; arXiv
Original source ↗Archived evidence ↗
Optimizing Visibility in Generative Engines: A Critical Survey of Generative Engine Optimization (2023–2026)
arXiv
Original source ↗Archived evidence ↗
Bringing Agentic Search to Earth Observation Data Discovery
arXiv; accepted at CIKM 2026
Original source ↗Archived evidence ↗
An Eye Tracking Study: Are AI Overviews Changing Search Behavior?
ACM SIGIR 2026; Microsoft Research
Original source ↗Archived evidence ↗
Announcing the Monetization Gateway: charge for any resource behind Cloudflare via x402
Cloudflare
Original source ↗Archived evidence ↗
Content Independence Day, one year on: building the business model for the agentic Internet
Cloudflare
Original source ↗Archived evidence ↗
LoRE: Enhancing Search Relevance with Progressive Chain-of-Thought and Preference Alignment
Association for Computational Linguistics; Taobao & Tmall Group (Alibaba)
Original source ↗Archived evidence ↗
Announcing the Agentic Resource Discovery specification
Google Developers Blog
Original source ↗Archived evidence ↗
Beyond Monolingual Deep Research: Evaluating Agents and Retrievers with Cross-Lingual BrowseComp-Plus
arXiv
Original source ↗Archived evidence ↗
One Polluted Page Is Enough: Evaluating Web Content Pollution in LLM Recommenders
arXiv
Original source ↗Archived evidence ↗
SafeGEO: Understanding Generative Engine Optimization Risks in Recommendation Agents
arXiv
Original source ↗Archived evidence ↗
Rethinking Literature Search Evaluation: Deep Research Helps, and Human Citation Lists Are Not a Ground Truth
arXiv; Mila / HEC Montréal and collaborators
Original source ↗Archived evidence ↗
Information Diversity and Authority Shifts in Google’s AI Overviews: An Audit of Health Queries
Association for Computing Machinery (ACM), WebSci Companion 2026
Original source ↗Archived evidence ↗
Retrieval as Reasoning: Self-Evolving Agent-Native Retrieval via LLM-Wiki
arXiv; Tencent WeChat research
Original source ↗Archived evidence ↗
What Gets Cited: Competitive GEO in AI Answer Engines
arXiv; ACM SIGIR 2026
Original source ↗Archived evidence ↗
Deep-Research Agents Can Be Poisoned via User-Generated Content
arXiv / Cornell Tech
Original source ↗Archived evidence ↗
SGR-Bench: Benchmarking Search Agents on State-Gated Retrieval
arXiv
Original source ↗Archived evidence ↗
A new resource for optimizing for generative AI in Google Search
Google Search Central
Original source ↗Archived evidence ↗
Optimizing your website for generative AI features on Google Search
Google Search Central
Original source ↗The Impact of AI Search on the Online Content Ecosystem: Evidence from Google and Reddit
arXiv
Original source ↗Archived evidence ↗
Measuring Google AI Overviews: Activation, Source Quality, Claim Fidelity, and Publisher Impact
arXiv (original preprint)
Original source ↗Archived evidence ↗
Superintelligent Retrieval Agent: The Next Frontier of Information Retrieval
arXiv
Original source ↗Archived evidence ↗
5 new ways to explore the web with generative AI in Search
Archived evidence ↗
Beyond Semantic Similarity: Rethinking Retrieval for Agentic Search via Direct Corpus Interaction
arXiv
Original source ↗Archived evidence ↗
How Generative AI Disrupts Search: An Empirical Study of Google Search, Gemini, and AI Overviews
Original research authors / arXiv
Original source ↗Archived evidence ↗
From Citation Selection to Citation Absorption: A Measurement Framework for Generative Engine Optimization Across AI Search Platforms
arXiv
Original source ↗Archived evidence ↗
The attribution crisis in LLM search results: Estimating ecosystem exploitation
Cambridge University Press, Data & Policy
Original source ↗Archived evidence ↗
AgentSearchBench: A Benchmark for AI Agent Search in the Wild
arXiv
Original source ↗Archived evidence ↗
From Experience to Skill: Multi-Agent Generative Engine Optimization via Reusable Strategy Learning
arXiv / original research authors
Original source ↗Archived evidence ↗
Think Before Writing: Feature-Level Multi-Objective Optimization for Generative Citation Visibility (original preprint)
arXiv
Original source ↗Archived evidence ↗
CoSearch: Joint Training of Reasoning and Document Ranking via Reinforcement Learning for Agentic Search
arXiv (UMass Amherst / Snap Inc. research)
Original source ↗Archived evidence ↗
MERRIN: A Benchmark for Multimodal Evidence Retrieval and Reasoning in Noisy Web Environments
arXiv
Original source ↗Archived evidence ↗
Retrieval Collapses When AI Pollutes the Web
ACM Web Conference 2026 (WWW '26) proceedings
Original source ↗Archived evidence ↗
Don't Measure Once: Measuring Visibility in AI Search (GEO)
arXiv
Original source ↗Archived evidence ↗
New ways Google is using AI to make shopping easier
Google India Blog
Original source ↗Archived evidence ↗
The Impact of Google AI Overviews on Publisher Traffic and User Experience: Evidence from a Field Experiment
SSRN (Indian School of Business; Carnegie Mellon University)
Original source ↗Archived evidence ↗
Inside Googlebot: demystifying crawling, fetching, and the bytes we process
Google Search Central
Original source ↗Archived evidence ↗
Structural Feature Engineering for Generative Engine Optimization: How Content Structure Shapes Citation Behavior
arXiv
Original source ↗Archived evidence ↗
Shopping with a Platform AI Assistant: Who Adopts, When in the Journey, and What For
arXiv; Ctrip collaborative empirical research
Original source ↗Archived evidence ↗
The End of Rented Discovery: How AI Search Redistributes Power Between Hotels and Intermediaries
Blossom AI / arXiv
Original source ↗Archived evidence ↗
Answer Bubbles: Information Exposure in AI-Mediated Search
arXiv; University of Illinois Urbana-Champaign
Original source ↗Archived evidence ↗
Shopping Companion: A Memory-Augmented LLM Agent for Real-World E-Commerce Tasks
Alibaba International Digital Commercial Group / arXiv
Original source ↗Archived evidence ↗
Structured Linked Data as a Memory Layer for Agent-Orchestrated Retrieval
Original research authors / arXiv
Original source ↗Archived evidence ↗
Diagnosing and Repairing Citation Failures in Generative Engine Optimization
arXiv
Original source ↗Archived evidence ↗
Quantifying Uncertainty in AI Visibility: A Statistical Framework for Generative Search Measurement
arXiv
Original source ↗Archived evidence ↗
AgenticGEO: A Self-Evolving Agentic System for Generative Engine Optimization
arXiv
Original source ↗Archived evidence ↗
MC-Search: Evaluating and Enhancing Multimodal Agentic Search with Structured Long Reasoning Chains
arXiv; original multimodal retrieval research
Original source ↗Archived evidence ↗
ProductResearch: Training E-Commerce Deep Research Agents via Multi-Agent Synthetic Trajectory Distillation
arXiv; Alibaba International Digital Commercial Group
Original source ↗Archived evidence ↗
SAGEO Arena: A Realistic Environment for Evaluating Search-Augmented Generative Engine Optimization
ACM SIGKDD 2026 / arXiv
Original source ↗Archived evidence ↗
Introducing AI Performance in Bing Webmaster Tools Public Preview
Microsoft Bing Blog
Original source ↗Archived evidence ↗
SAGE: Benchmarking and Improving Retrieval for Deep Research Agents
arXiv
Original source ↗Archived evidence ↗
Generative Engine Optimization: A VLM and Agent Framework for Pinterest Acquisition Growth
arXiv
Original source ↗Archived evidence ↗
Agentic Search in the Wild: Intents and Trajectory Dynamics from 14M+ Real Search Requests
arXiv
Original source ↗Archived evidence ↗
Agentic-R: Learning to Retrieve for Agentic Search
arXiv; Renmin University of China and Baidu Inc. affiliations
Original source ↗Archived evidence ↗
E-GEO: A Testbed for Generative Engine Optimization in E-Commerce
arXiv
Original source ↗Archived evidence ↗
Bing Introduces Support for the data-nosnippet HTML Attribute
Microsoft Bing
Original source ↗Archived evidence ↗
When Content is Goliath and Algorithm is David: The Style and Semantic Effects of Generative Search Engine
arXiv (original preprint)
Original source ↗Archived evidence ↗
DeepTRACE: Auditing Deep Research AI Systems for Tracking Reliability Across Citations and Evidence
arXiv
Original source ↗Keeping Content Discoverable with Sitemaps in AI Powered Search
Microsoft Bing
Original source ↗Archived evidence ↗
News Source Citing Patterns in AI Search Systems
arXiv original research preprint
Original source ↗Archived evidence ↗
Consent in Crisis: The Rapid Decline of the AI Data Commons
arXiv / NeurIPS 2024
Original source ↗Archived evidence ↗
Enabling Large Language Models to Generate Text with Citations
arXiv
Original source ↗Archived evidence ↗
WebGPT: Improving the factual accuracy of language models through web browsing
OpenAI
Original source ↗Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks
arXiv / Meta AI researchers
Original source ↗Answering customer questions with your knowledge base
Shopify Help Center
Original source ↗Archived evidence ↗
Auditing Citation Behavior in AI-Generated Search Summaries: A Framework and a Case Study of Google AI Overviews
Proceedings of Machine Learning Research (PMLR); Canadian Conference on Artificial Intelligence
Original source ↗Archived evidence ↗
Consent in Crisis: The Rapid Decline of the AI Data Commons (NeurIPS 2024 proceedings)
NeurIPS 2024 Datasets and Benchmarks Track
Original source ↗Archived evidence ↗
DailyReport: An Open-ended Benchmark for Evaluating Search Agents on Daily Search Tasks
Association for Computational Linguistics (Findings of EMNLP 2026)
Original source ↗Archived evidence ↗
Evidence Laundering: Deep Research Agents Mistake Correlated Sources for Independent Evidence
Association for Computational Linguistics (EMNLP 2026)
Original source ↗Archived evidence ↗
How Do LLMs Cite? A Mechanistic Interpretation of Attribution in Retrieval-Augmented Generation
University of Amsterdam / arXiv
Original source ↗Archived evidence ↗
One Polluted Page Is Enough: Evaluating Web Content Pollution in LLM Recommenders
Association for Computational Linguistics
Original source ↗Archived evidence ↗
SafeGEO: Understanding Generative Engine Optimization Risks in Recommendation Agents
Association for Computational Linguistics
Original source ↗Archived evidence ↗
Seeing Through the MiRAGE: Evaluating Multimodal Retrieval Augmented Generation
Association for Computational Linguistics (EMNLP 2026)
Original source ↗Archived evidence ↗
Think Before Writing: Feature-Level Multi-Objective Optimization for Generative Citation Visibility
Association for Computational Linguistics
Original source ↗Archived evidence ↗