[REVISION HISTORY]

AI market share, reliability, and web traffic trends

Updated 2 times since CLSTR started tracking revisions of this situation.

What changed

2026-09-28 05:11 UTC → 2026-09-29 02:32 UTC · added ~~removed~~

Recent data indicates shifting market dynamics and reliability concerns within the artificial intelligence sector. According to a Comscore report, ChatGPT’s dominance in AI prompt volume fell from 70% to 50% during the first half of 2026. Concurrently, Google’s Gemini nearly doubled its share from 17% to 30%, while Claude grew from 2% to 11%. Reliability issues persist, particularly in specialized fields. A study by Saturn found that popular AI models provided incorrect answers to financial queries in an average of 57% of cases, including errors regarding UK pension rules. ~~Additionally,~~ Further research ~~suggests that AI recommendation engines may exhibit~~ by Product.ai highlights significant hurdles for consumer use; a ~~brand notoriety bias, favoring established brands over unknown ones even when~~ study testing major models like ChatGPT, Claude, Gemini, and Perplexity across 220 shopping scenarios found that 86% of queries resulted in “reproducible factual contradictions.” These included inconsistent data regarding product ~~characteristics are identical.~~ prices, specifications, and model names. Dakota Nunley, head of search products at Product.ai, stated that LLMs have not yet reached a practical level for end-to-end shopping due to an inability to provide accurate foundational information. By late 2026, the landscape of AI-driven web traffic has shifted significantly. While prompt volume shares varied, ChatGPT has emerged as the dominant force in directing traffic to websites, accounting for 95.1% of all web traffic from AI assistants, compared to just 2.4% for Gemini. This dominance saw a 36% increase in August 2026. This shift has led businesses to seek new ways to influence AI responses. Rather than direct model manipulation, influence is increasingly exerted through the sources AI consults during real-time searches. Research from Princeton suggests that including citations and statistics can increase visibility in AI-generated responses by up to 40%. However, major players are cracking down on deceptive practices; Google has classified attempts to manipulate generative AI responses as spam since August 2026, and Microsoft has identified 31 companies using deceptive ‘Summarize with AI’ buttons.

Versions

Only revisions since CLSTR began indexing content versions appear here. Select a version to see what changed compared to the one before it.