Listen: AI Search Citation Mining
Raw data dump from a citation mining pipeline demo featuring 60 prompts across AEO, AI marketing, AI optimization, AI SEO, and AIO using GPT-5 and Gemini.
Transcript
A new citation mining pipeline from Dejan AI is shedding light on how artificial intelligence engines credit their sources. In a recent demo, the system tracked how different models answer prompts about topics like AI marketing and Answer Engine Optimization, commonly known as AEO.
The test ran over one hundred operations to see where these engines get their facts, and the results show a big difference in how they behave. Google’s Gemini engine was highly active, generating four hundred citations, while OpenAI produced about one hundred and forty.
When backing up its answers, OpenAI relied heavily on Google’s developer documentation, Semrush, and Dejan AI's own research on multi-step research agents. Gemini took a different path, frequently citing digital marketing blogs and business resource sites like Single Grain.
By mapping out these citation networks, this pipeline reveals exactly which websites are shaping the answers generated by today's leading AI engines.