266 items, newest first. Page 3 of 12.
New benchmark reveals AI platforms can't reliably detect GEO-optimized content gaming LLM citations.
Claude's watermarking shifts word choices invisibly, creating content integrity risks for precision-language brands.
Competing GSEs face a collective action trap that structurally erodes the web corpus their answers depend on.
GSEs favour a narrow publisher set. Authoritative institutions outside it vanish from AI answers, regardless of content quality.
ChatGPT encodes brand preferences before searching. Named brands get a 33x visibility edge.
OpenAI's fetch bot bypasses robots.txt blocks, undermining brands' assumed control over AI retrieval.
Gemini 3.7 Flash hits AI Mode in 24 hours, concentrating citations and raising the bar for B2B content.
Star ratings move LLM recommendations by 31 points. Structured reputation signals now determine AI visibility.
GPT-5.6's routing logic means brand content must now survive a gatekeeping step before it reaches the reasoning model.
OpenAI's Ultrafast tier hits 750 tokens/sec via Cerebras. Here is what it changes for enterprise AI architectures.
Practitioners want GEO data but distrust the vendors selling it, per new Search Engine Journal survey.
Gemini 3.7 Flash halves cost in three weeks, intensifying the race to power enterprise agentic pipelines.
OpenAI opens ChatGPT to health and finance ads, raising stakes for organic LLM authority.
Grok 4.6 matches OpenAI's best model at 60% lower cost, threatening LLM citation strategies built around a single model.
Gemini's share fell from 12% to 1.9%. For brand AI-visibility strategy, that concentration risk is immediate.
ChatGPT cites small unlicensed sites as often as major licensed publishers, new data shows.
Anthropic's universal Claude watermark makes AI-drafted content detectable by platforms and LLMs from August 2026.
Anthropic's Opus 5 system prompt patches post-cutoff knowledge gaps with a plain-text correction about its own suspension.
Google dismantles DeepMind's autonomy and moves Gemini west as Hassabis prepares to exit.
The cats.txt test demolishes the evidence for llms.txt, exposing a wider GEO credibility problem.
OpenAI upgrades Sol's accuracy and moves Luna to free tier, creating two distinct citation environments for brands.
OpenAI's free-vs-paid split puts deeper reasoning in professional hands. B2B brands must respond.
Web search mode changes ChatGPT's citation behaviour systematically, an arXiv audit of 4,812 responses finds.
Google Search Console's new AI impression data exposes a growing split between ranking and AI citation.