GEO news
What is changing in AI search. Each item is a summary with a link to the outlet that reported it.
Anthropic cuts live internet access for all internal evaluations after Claude models reached real sites
Anthropic has extended its live internet cutoff to every internal evaluation after finding cases where Claude models interacted with external websites in ways it did not intend. The company says the affected sites include pages run by federal, state and local U.S.
Source: Search Engine Journal
Anthropic bans using Claude to seed fake sources for AI answers
Anthropic's updated usage policy, effective 12 November, bars people from using Claude to seed the sources that search engines and AI systems draw answers from with content that misrepresents its origin, authorship or independence.
Source: Search Engine Journal
Karpathy: Ask Models for ASD-STE100 Controlled English, Diagrams, HTML and Explainer Videos
Andrej Karpathy, a founding member of OpenAI who joined Anthropic's pretraining team in May, said on X that asking language models to explain things in ASD-STE100 produces more readable output.
Source: Search Engine Journal
69% of Marketers Are Publishing More AI Content Than Last Year (New Research Report)
Ahrefs has published its second annual survey of marketers on how AI is changing content production. Of 301 respondents, 280 (93%) said they use AI to create or help create content, up from 87% a year earlier. Among the small group who avoid it, accuracy was the main concern.
Source: Ahrefs Blog
Braze expands AI from content creation to campaign operations
Braze announced three new AI tools at its Forge conference in Las Vegas: BrazeAI Decisioning Studio Go, Agentic Standards, and Operator Connect.
Source: MarTech
Anthropic tells developers to retune prompts for Claude Opus 5.5
Anthropic has published prompting guidance for Claude Opus 5.5, the model it released on 22 September, and much of it asks developers to revisit settings carried over from Opus 5.
Source: Search Engine Journal
Claude and Claude Code Show Different Search Behaviours, Data Finds
Claude and Claude Code behave differently when answering the same prompts, according to new data from Profound, which monitors how brands appear in AI answers.
Source: Search Engine Journal
