One brand's AI visibility score ranged from 15.5% to 59.5% across five AI engines — running on the same set of queries. A 900-prompt study published September 25, 2026 by Zen Media documented this 44-point spread across ChatGPT, Gemini, Claude, Perplexity, and Grok. The variance is structural: each AI platform retrieves and cites brands through a different architecture, pulling from different source pools. BrandCited monitors all eight major AI engines — ChatGPT, Perplexity, Claude, Gemini, Copilot, Grok, You.com, and Brave — and returns a per-engine breakdown alongside the composite score. Brands tracking only one engine are missing the gaps on the other seven.
BrandCited's free AI visibility audit checks all eight engines and shows your score in 30 seconds. Run it at brandcited.ai.
What did the September 2026 AI visibility study find?#
One brand's coverage ranged from 15.5% to 59.5% across five AI engines when tested against 900 identical prompts. The Zen Media study, published via GlobeNewswire on September 25, 2026, tracked branded and non-branded discovery together, finding that 85% to 87% of prompts that surfaced the client did not mention the company by name.
Atomic fact: 85% to 87% of the prompts that surfaced a brand across AI engines in the Zen Media study were non-branded queries — the user never typed the company name.
Atomic fact: A single brand tested across five AI engines showed a 44-point spread between its highest visibility score (59.5% on one engine) and its lowest (15.5% on another), running on the same 900 prompts.
The study also found that brands earning both a mention signal and a citation signal showed a 40% higher likelihood of reappearing across subsequent answers. Only 28% of AI answers in the study included brands with that dual visibility.
Why do ChatGPT, Perplexity, and Gemini cite such different brands?#
Each AI engine retrieves content through a different architecture, so each engine draws from a different source pool. Across 680 million AI citations analyzed in early 2026, only 14% of top domains were cited across all four of ChatGPT, Claude, Gemini, and Perplexity. Only 11% of domains appeared in both ChatGPT and Perplexity citations.
Atomic fact: Only 11% of domains are cited by both ChatGPT and Perplexity, based on analysis of 680 million AI citations from early 2026.
Atomic fact: Only 14% of top domains appear across all four of ChatGPT, Claude, Gemini, and Perplexity — meaning 86% of what one engine cites is invisible to the other three.
ChatGPT uses Bing search and its own training data. Perplexity runs a fresh web search on every query. Claude (Anthropic) weights semantic coherence and entity graph linkouts. Gemini grounds citations in Google's Knowledge Graph first. These four architectures reward different content attributes, which explains why a brand that leads one engine's citation pool can be absent from another's.
Search Engine Journal's AI visibility data for 2026 shows ChatGPT at 53.9% of worldwide AI assistant web traffic, Gemini at 27.9%, and Claude at 9.2%.
How did GPT-6 change ChatGPT's citation behavior?#
GPT-6 Astra, released September 3, 2026, and GPT-6 Sol and Luna, released September 22, 2026, shifted how ChatGPT retrieves and cites brand content. Astra expanded ChatGPT's browsing capability — the model retrieves more live web sources per query than GPT-5 and applies stronger source-quality filtering.
Atomic fact: GPT-6 Sol makes roughly half as many factual mistakes as its predecessor, approaching GPT-6 Astra-level reliability at lower cost — ChatGPT citations now come from a more quality-filtered source pool than six months ago.
Atomic fact: GPT-6 Sol and Luna became available to all ChatGPT Plus, Pro, Business, Enterprise, and Edu users as of September 22, 2026, meaning the citation behavior change reached the entire paying ChatGPT user base within 24 hours of launch.
For brands whose content previously surfaced in ChatGPT via training-data memory, the GPT-6 shift is a citation reset. Brands without fresh, structured, citable web content will see their ChatGPT scores drop as the new models become the default.
What does a 44-point gap between your best and worst AI engine actually mean in practice?#
A 44-point spread in AI visibility across engines translates to a measurable customer acquisition gap. Google AI Overviews now appear on 47% to 64% of all queries in 2026. Google reported 2.5 billion monthly active users for AI Overviews and more than 1 billion for AI Mode as of mid-2026. ChatGPT serves 900 million users a week.
Atomic fact: Queries triggering Google AI Overviews show a 30% to 50% average reduction in organic click-through rate for the top three positions — a brand's Google ranking generates less traffic than two years ago even when the ranking holds.
Atomic fact: Research from Brandlight shows the overlap between the top Google organic results and the sources AI engines cite dropped from 70% in 2024 to below 20% in 2026.
Which AI engines drive the most brand discovery in 2026?#
ChatGPT drives the most raw brand discovery by volume at 53.9% of AI assistant web traffic, but Gemini and Perplexity outperform on specific discovery paths. Airops's 2026 State of AI Search found that most AI-driven brand discoveries happen through non-branded category queries.
Atomic fact: ChatGPT's share of AI assistant web traffic fell from 76.4% in mid-2025 to 53.9% in May 2026 as Gemini rose from 9% to 27.9%.
Atomic fact: Gemini reached 1 billion monthly users as of August 2026, making it the second-largest AI assistant by user base.
What AI search updates happened in the last 24 hours?#
Track your AI visibility for free
See how ChatGPT, Claude, Gemini, and 4 other AI platforms mention your brand.
- OpenAI ChatGPT Ads expansion: OpenAI expanded ChatGPT Ads to seven more Asian markets, bringing availability to more than 60 countries. ChatGPT Ads reached $1 billion in annualized revenue in under 200 days after launch.
- Perplexity + Microsoft 365 rollout: Perplexity rolled out Computer as a native add-in inside Word, Excel, PowerPoint, and Outlook for all Pro, Max, and Enterprise subscribers. Brands absent from Perplexity's citation pool are now invisible inside millions of enterprise users' daily workflows.
- GPT-6 Sol stable release: ChatGPT's stable release on September 25, powered by GPT-6 Sol, completed rollout to all Plus and Pro users, finalizing the shift from GPT-5 as the default model.
- Google AI Overviews expansion: September 2026 data confirms AI Overviews cover 47% to 64% of all queries.
- GEO adoption accelerates: Search Engine Land's 2026 GEO guide reports that the majority of SEO teams now have a GEO workstream alongside traditional SEO.
How does BrandCited audit per-engine AI visibility?#
BrandCited runs each brand's query set against all eight AI engines separately and returns a per-engine score alongside the composite. The per-engine breakdown shows exactly which queries each engine surfaces your brand for, which it misses, and what the competing brands look like on each engine. Each engine's citation behavior is logged as a separate audit trail so score changes can be attributed to specific model updates rather than appearing as an unexplained composite drop.
See your brand's per-engine breakdown in BrandCited's dashboard. Run a free audit at brandcited.ai — no signup required.
What should you do right now?#
- 1Run a per-engine AI visibility audit today. The 44-point variance found in the Zen Media study is not an edge case. Check your brand at brandcited.ai and find which engines are gaps versus which are performing.
- 1Publish at least two structured articles per week. GPT-6 Astra's retrieval layer weights recent live-web content over training-data memory. Brands with content last published six or more months ago are at a citation disadvantage on ChatGPT compared to brands publishing actively.
- 1Add FAQ schema to your top five pages. BrandCited's scan data across 2,000+ tracked brands shows sites with complete
FAQPage JSON-LD are cited more often in Perplexity answers than comparable sites without it.
- 1Submit your sitemap to Bing Webmaster Tools. ChatGPT uses Bing for its live-retrieval layer. Brands not indexed in Bing are invisible to ChatGPT's web retrieval regardless of their Google rankings.
- 1Verify your Google Business Profile and request a Knowledge Panel. Gemini grounds citations in the Google Knowledge Graph before pulling fresh retrieval. A verified Knowledge Panel raises Gemini's citation confidence for category queries.
- 1Track your score weekly during model rollout periods. GPT-6 Sol and Luna shipped to all paid ChatGPT users on September 22. Citation scores can shift 10 to 20 points within two to three weeks of a major model rollout. Set up BrandCited's score-drop alerts so a change appears in your inbox the day it happens.
Frequently asked questions#
Why do different AI engines show different results for the same brand?
Each AI engine retrieves content through a different architecture. ChatGPT uses Bing search and its own training data. Perplexity runs fresh web searches. Claude weights semantic coherence and entity graphs. Gemini grounds citations in the Google Knowledge Graph first. A brand visible in ChatGPT answers can be absent from Perplexity answers for the same query.
What is a good AI visibility score?
BrandCited's dataset of 2,000+ tracked brands shows the top-quartile threshold is a composite score of 70 out of 100. The median is 38. Category leaders in competitive B2B SaaS score between 75 and 92. A brand below 20 is functionally invisible to AI-driven discovery.
How many AI engines should I monitor for brand visibility?
Monitor at least five engines: ChatGPT, Perplexity, Gemini, Claude, and Grok. Only 14% of top domains overlap across the four biggest engines, so monitoring one engine misses the citation gaps on the other seven.
Does ChatGPT cite the same brands as Perplexity?
No. Across 680 million AI citations analyzed in early 2026, only 11% of domains appeared in both ChatGPT and Perplexity citations. The two engines' brand citation pools are almost entirely disjoint.
What did GPT-6 change about ChatGPT's citation behavior?
GPT-6 Astra (September 3, 2026) and GPT-6 Sol and Luna (September 22, 2026) expanded ChatGPT's browsing capabilities. The new models retrieve more live web sources per query and apply stronger quality filtering, meaning brands without fresh structured web content will see their ChatGPT visibility drop as GPT-6 becomes the default model.
How often should I check my AI visibility score?
Check monthly at minimum. Engine citation behavior changes with every model update. BrandCited runs automated scans and alerts you when your score drops more than 10 points on any single engine.
Sources#
- 1Zen Media (2026). One Brand's AI Visibility Tracking Ranged From 15.5% to 59.5% Depending on the AI Engine. GlobeNewswire.
- 2OpenAI (2026). GPT-6 Astra: A new generation of intelligence. OpenAI Blog.
- 3OpenAI (2026). Introducing GPT-6 Sol and Luna. OpenAI Blog.
- 4TechCrunch (2026). OpenAI launches GPT-6 Sol and Luna. TechCrunch.
- 5Search Engine Journal (2026). ChatGPT, Gemini & Claude Lead AI Visibility. SEJ.
- 6QuickSEO (2026). ChatGPT vs Perplexity for AI Visibility in 2026. QuickSEO.
- 7Peec.ai (2026). . Peec AI.
