AI Answers Lean on People but Cite Platforms, iSonic Test Finds

In an iSonic test of 1,728 answers from four AI assistants, four of the five most cited websites were Reddit, LinkedIn, Medium and YouTube, places where people share what they know. Founder Ankesh Kumar calls it a case for Authentic Intelligence.

Palo Alto, California – iSonic, a platform that helps creators get found and credited by AI assistants, tested 1,728 answers from ChatGPT, Claude, Gemini and Perplexity. When people ask an AI assistant a question, much of the answer comes from other people, but what comes back is a summary with a list of websites underneath.

Four of the five websites cited most often were Reddit, LinkedIn, Medium and YouTube, platforms built on people sharing what they know. In the answers, that knowledge typically appeared as an overview, with the source reduced to a site name in a list.

The test ran on September 9, 2026. iSonic put the same 432 questions about AI, search and creator work to each of the four assistants and recorded the sources cited in every answer. Reddit appeared in 284 answers (16.4 percent), LinkedIn in 223 (12.9 percent), Medium in 123 (7.1 percent) and YouTube in 119 (6.9 percent). The trade publication Search Engine Land rounded out the top five at 73 answers. The rest of the top ten were company and institutional sources: Semrush (67), iSonic’s own site (58), Google’s developer documentation and the research archive arXiv (51 each), and Ahrefs (49). iSonic’s site appeared because 38 of the 432 questions named a brand.

The assistants behaved very differently. Perplexity accounted for 230 of the Reddit citations and 216 of the LinkedIn ones, while ChatGPT drew mostly on official documentation. The results describe one day and one topic area. The counts, per-assistant results and method are published in iSonic’s write-up of how the test was run.

“The problem with artificial intelligence is the word artificial,” said Ankesh Kumar, Founder and CEO of iSonic. “What we believe in is authentic. When someone asks a question, they want the answer from a person who has lived it. Today they get a summary and a list of links. The real people are in there. You just never hear from them.”

The findings come amid rising public unease about AI. In a Pew Research Center survey of 5,023 U.S. adults conducted June 9 to 15, 2025, 50 percent said they were more concerned than excited about the growing use of AI in daily life, up from 37 percent in 2021. Half said AI will make it harder for people to form meaningful relationships. Just 5 percent said it will make that easier.

“Authentic intelligence should lean into something very human, which is the desire for connection,” Kumar said. “Asking a question and hearing back from a real person shouldn’t be one more thing we lose. If an AI learned something from you, it should be able to say your name.”

For creators, the gap is practical. A recipe developer, a travel photographer or a parenting writer may have years of answers spread across posts, videos and podcasts. Before an assistant can point back to the person behind an answer, it has to be able to read the work and tell who wrote it. When that signal is missing, the platform gets the credit by default.

The gap is not only about credit. Some of the platforms at the top of the list have been paid for their content. Reddit, the most cited site in the test, has signed data licensing deals with Google and OpenAI. The people who wrote the posts behind those answers typically receive neither payment nor a name in the answer. As AI companies and publishers negotiate over who should be paid for training data and answers, individual creators are largely absent from the conversation.

“Before anyone can talk about paying creators, AI has to know who they are,” Kumar said. “Right now the platform gets the deal and the AI gets the answer. The person who did the work gets nothing, not even their name. Credit is the first step, and it’s the one we can fix today.”

About iSonic

iSonic is a GEO (generative engine optimization) platform that turns a creator’s existing content into structured, AI-readable pages so AI assistants can find it and credit it to the creator. Creators bring their posts, videos, podcasts and links, and iSonic builds an AI Bio Page that assistants such as ChatGPT, Claude, Gemini and Perplexity can read. iSonic is based in Palo Alto, California. More about how it works is on the iSonic features page.

Contact Information

Business: iSonic

Contact Name: Ankesh Kumar, Founder and CEO 

Email: [email protected] 

Website: https://isonic.ai/  

Country: United States

Disclaimer: This article is for informational purposes only. The findings attributed to iSonic reflect the study’s results and may not represent all AI platforms or user experiences. Readers should independently verify the information before drawing conclusions.