New data shows 40% of AI citations omit the source brand entirely. On Perplexity, it hits 52%. Here's what that means for your visibility and how to protect it.
40% of AI citations across 16 million brand appearances fail to name the source brand
Getting cited by an AI search engine sounds like a win. Your page made it into the answer. Your URL shows up in the sources. The engine found your content useful. But a growing body of analysis reveals a hidden cost: AI systems are stripping your brand name from the citation while keeping your content.
52% of citations on Perplexity omit the source brand name entirely
When a customer gets an answer from an AI search engine, they see the generated text first. Most never scroll down to check the sources. If your brand name does not appear in the visible answer, you get zero credit for the work. Your URL might be sitting in a collapsed sources list, but your brand, your expertise, and your authority stay invisible.
This breaks the entire value chain of content marketing. You invested in research, writing, and optimization. AI picked it as authoritative enough to answer a user question. But the user walks away having no idea where that answer came from. No brand recall. No trust buildup. No traffic.
Perplexity's 52% ghost citation rate is especially significant because the engine is one of the fastest-growing AI search platforms. If your industry competitors publish on topics Perplexity regularly answers, and half the time your brand gets anonymized while theirs does not, you are losing visibility at scale. The citation omissions compound over time across thousands of user queries.
Ghost citations are not the only structural issue worth understanding. A recent developer post from DEV Architecture highlights something most business owners never see: the AI answers being served to your customers are often generated by systems juggling four or more separate LLM providers at once, each with different rate limits, throttling rules, and onboarding requirements. OpenAI, Anthropic, Gemini, and DeepSeek all behave differently under load.
Why does that matter to you? Because when an AI search engine degrades in quality due to congestion or quota limits, the answers it produces about your industry get shorter, vaguer, and less likely to name any specific source brand. The ghost citation problem gets worse, not better, when the underlying infrastructure is strained. Your content may get deprioritized or truncated in exactly the moments when query volume is highest.
The ghost citation problem is still evolving, and AI search engines are still refining how they attribute sources. But the trend is clear: being cited is no longer enough. Being cited by name is what builds your brand and drives your business. If you are not tracking which of your citations are ghosts, you are flying blind on a critical part of your AI search visibility.
A ghost citation occurs when an AI search engine uses your content to generate an answer but fails to name your brand in the response, even though it may link to your URL in a sources list. Your work is doing the heavy lifting, but customers don't know where it came from.
Most users read the AI-generated answer and never scroll to see the sources list. Without your brand name in the visible answer, you lose brand recognition, trust signals, and click-through traffic, even though AI confirmed your content was valuable enough to cite.
Perplexity has the highest documented rate at 52% of citations omitting the brand name, according to Writesonic's analysis of 16 million brand appearances. Other engines show rates around 40% on average.
The source doesn't detail specific tactics, but the immediate options are: negotiate better terms with AI engines about attribution, monitor where your content appears and request corrections, or be selective about which high-value content you allow AI systems to index.