Search Visibility OS
OpenAI launches ChatGPT Search
OpenAI launched ChatGPT Search with timely web answers and links to relevant sources.
Controlled research finds location channels change AI citations and business recommendations
Chen and Koudas report a 31,200-answer controlled audit of how declared and network location influence sources cited and businesses recommended by tested ChatGPT and Gemini configurations.
SEEDS research separates retrieved candidates from evidence admitted into agent reasoning
Liu and colleagues introduced SEEDS, a preview-guided evidence admission framework for multi-round search agents. The agent selects which retrieved candidates to inspect in full rather than automatically reading a top-ranked prefix. The authors evaluated nine question-answering benchmarks across local and live-web search and reported accuracy improvements with fewer full-document reads.
OpenAI introduces Sponsored Agents in ChatGPT Ads
OpenAI announced Sponsored Agents, an advertising format that lets a user start a clearly labeled conversation with a business-sponsored agent after clicking an ad in ChatGPT.
Research traces ChatGPT Web retrieval-to-citation behavior across model settings
An original observational study compares four ChatGPT Web settings from GPT-5.2 to GPT-5.5, examining query rewriting, retrieved candidates and cited sources at URL, hostname and entity levels, with external Google Search and OpenAI Responses API proxies.
Controlled research finds source-name preferences in LLM agent item selection
Song and colleagues evaluated 12 LLM agent models across shopping, accommodation and scholarly search, finding that the source associated with an item can influence selection even after controlling item requirements and position.
RAID study tests GEO rewriting without access to future search queries
An original preprint evaluates RAID, a role-augmented intent-driven rewriting method that infers likely search needs from source content without seeing future queries. Its simulated generative-search benchmark reports modest average improvements across 500 held-out query instances.
ReCite evaluates intent-aware retrieval and verification for faithful academic citations
Huang and colleagues introduce ReCite, a peer-reviewed agentic academic citation system that identifies citation locations, plans searches from citation intent, verifies claim-evidence support and retries retrieval after mismatches.
SAGE benchmark finds agent query wording changes retrieval effectiveness
SAGE introduces 1,200 scientific-literature questions across four domains and approximately 200,000 papers. In a fixed-corpus DR Tulu comparison, lexical BM25 surpassed tested semantic retrievers on short-form exact match, while a semantic retriever led on open-ended weighted recall.
EMBL AI Librarian demonstrates multi-query evidence retrieval in scientific research agents
Sigillo and colleagues introduce EMBL AI Librarian, a life-sciences knowledge layer that plans complementary subqueries over live Europe PMC search, reads selected papers and locates evidence for grounded answers. Their original preprint reports improvements on citation and question-answering benchmarks.
SearchHorizonBench separates answer success from the marginal value of another agent search
Ning, Li and Kong introduce SearchHorizonBench, a paired-continuation evaluation of SEARCH versus THINK from the same agent state, with 45,632 executions across 3,168 questions. The study finds that calibration of overall success can improve while predicting the incremental benefit of search remains difficult.
ITER uses prior agent search history to improve iterative evidence retrieval
Chen and colleagues introduced ITER, a dense retriever trained with agent interaction trajectories that incorporates the main question, earlier subqueries and pre-search reasoning when retrieving evidence for subsequent search steps.
ACARe research shows task context can improve reranking of agent-generated search queries
An original preprint introduces a 4B agent-context-aware reranker using the user's original question and available agent reasoning context alongside generated subqueries. In the reported controlled benchmarks it improves dataset-averaged nDCG@10 by 0.9–2.0 percentage points over a base reranker and reduces searches for five tested agents.
Research separates citation selection from citation absorption in AI search
A 2026 preprint proposes a two-stage GEO measurement framework separating citation selection from citation absorption and evaluates 602 controlled prompts across ChatGPT, Google AI Overview/Gemini and Perplexity using a public dataset.
LLM-Wiki research shows agent-guided traversal of linked evidence can improve multi-hop retrieval
Tencent-affiliated researchers introduce LLM-Wiki, an experimental agent-native retrieval system that compiles documents into interlinked Wiki pages and lets agents search, read and traverse relationships while checking evidence sufficiency. On three sampled multi-hop QA benchmarks, it reports higher F1 than tested flat and graph-enhanced retrieval baselines.
Mechanistic study separates citation-marker generation from evidence use in RAG
van Dort and Heuss use activation patching in Llama-3.1-8B-Instruct to identify distributed components involved in deciding whether to emit inline citations, then test interventions on PopQA-derived prompts and HotpotQA.
SIRA demonstrates corpus-aware one-shot retrieval as an alternative to repeated agent search
The SIRA research preprint combines LLM-proposed evidence vocabulary, corpus-frequency checks and a single weighted BM25 retrieval action. It reports competitive or stronger results than selected dense and multi-round agentic retrieval baselines in controlled benchmarks.
T-Search research separates multi-step evidence retrieval from answer generation
Tsymboi and colleagues introduce T-Search, an open-weight multi-round agentic retriever over fixed corpora that returns ranked evidence chunks and justifications to a separate answer generator. Across seven English and Russian benchmarks with gold evidence labels, the authors report mean Recall@10 of 56.0 in one rollout, 14.4 points above their base model, and 61.3 with three fused rollouts.
AgentGEO research diagnoses and repairs citation failures across simulated generative-engine stages
An original research preprint introduces a citation-failure taxonomy, an iterative AgentGEO diagnosis-and-repair framework, and the MIMIQ document-centric benchmark with 204 webpages and 12,240 generated queries. In controlled simulated generative-engine evaluation, the authors report higher citation rates after targeted repair.
IndexAct research separates candidate-set refinement from text inspection in agentic search
An original work-in-progress preprint introduces IndexAct, an index-native interface allowing search agents to manipulate persistent candidate sets using lexical conditions and set operations while retrieving text passages separately. The authors report improvements over evaluated baselines on five benchmarks.
Research benchmarks citation-ranking models across ten LLM engines under controlled retrieval
Bajemon and Rochet benchmarked citation allocation on 229 GEO-Bench queries across ten LLM engine configurations given five fixed passages per query, reporting stronger predictions from comparative decision models than from generic relevance rerankers and releasing a distilled 4B ranker.
XBCP research finds language mismatch weakens agentic retrieval and citation fidelity
Lu and colleagues introduce Cross-lingual BrowseComp-Plus (XBCP), keeping questions and answers in English while varying the languages of supporting evidence. They evaluate four deep-research agents and sparse/dense multilingual retrievers; the multilingual condition distributes documents across 12 languages. Authors report lower evidence recall, answer accuracy and citation reliability with language-mismatched evidence.
MAGEO research tests reusable GEO strategies with fidelity-aware citation evaluation
An original research preprint introduces MAGEO, a multi-agent GEO framework that learns reusable, engine-specific content-editing strategies. It also introduces twin-branch evaluation with a frozen retrieval list and DSV-CF metrics combining semantic visibility with attribution fidelity.
Research characterizes cross-platform conversational web search behavior
A research preprint studies the end-to-end web-search lifecycle across ChatGPT, Claude, Grok, and DeepSeek using real-world interactions and controlled experiments.
Cross-engine GEO audit finds little exact-URL citation overlap in a fixed-prompt sample
Benjamin Tannenbaum reports a June 6, 2026 observational audit of 15 fixed commercial prompts across ChatGPT, Microsoft Copilot, Google and Perplexity. The study reports 589 citation observations representing 528 distinct URLs and 356 domains; mean pairwise same-prompt exact-URL Jaccard overlap was 0.0079, and 84.9% of compared engine pairs shared no cited URL.
Research finds AI-search source selection and answer synthesis diverge across systems
A March 2026 original preprint compared 11,000 real-world information queries across GPT-4o-mini API configurations, Google AI Overviews and Google organic results. It reported low overlap between top cited domains of the OpenAI API web-search condition and Google, and uneven use of cited source content in generated answers.
OpenAI details agentic retrieval, ranking and auction architecture for ChatGPT Ads
OpenAI's engineering account describes a separate ChatGPT Ads service using a post-trained LLM orchestrator to interpret evolving conversation intent, generate semantic retrieval queries, retrieve and filter candidates, and coordinate ranking and auction tools.
ChatGPT Search
Web search and source-linked answers inside ChatGPT.
In a controlled 31,200-answer audit, Chen and Koudas observed different citation-source and business-recommendation outcomes across locations; the tested ChatGPT and Gemini configurations used different declared-versus-network location signals.
Evidence: emerging
In a study using 11,000 Natural Questions queries, Huang and colleagues reported that the GPT-4o-mini API web_search condition shared 25% of its top 100 cited domains with Google organic results and 24% with Google AI Overviews; the study also observed uneven representation of cited sources within synthesized answers.
Evidence: emerging
ChatGPT Ads expands to 26 new countries as over 100,000 advertisers have used the platform
OpenAI · official
Developers can now submit apps to ChatGPT
OpenAI · official
ChatGPT release notes
OpenAI · official