Google’s dominance in the search business has been under pressure ever since OpenAI launched ChatGPT in late 2022. The competition grew sharper last year when OpenAI added search functionality to ChatGPT, putting it head-to-head with Google.
Microsoft’s Bing and startups like Perplexity were already in the AI search race, but ChatGPT’s search was seen as so strong that OpenAI CEO Sam Altman openly admitted:
“I legitimately cannot tell you the last time I did a Google search.”
The Allegations
According to a new report by The Information:
OpenAI has been relying on data scraped from Google Search to improve ChatGPT’s real-time answers.
This is especially true in fast-changing areas like news, sports, and financial markets, where ChatGPT’s training data is weaker.
The data was accessed through SerpApi, a paid service that extracts live Google search results.
SerpApi’s clients also include Meta, Apple, and Perplexity. While it once listed OpenAI as a client, that mention has since been removed.
Evidence From Researchers
Former Google engineer Abhishek Iyer showed that ChatGPT was pulling information that only existed inside Google’s index. He created fake pages visible solely in Google Search, and ChatGPT was still able to use that data.
Google’s Response
OpenAI has always said ChatGPT relies on its own web crawler, Microsoft Bing, and content partnerships.
But during Google’s antitrust trial in 2024, OpenAI executive Nick Turley admitted that:
OpenAI had asked Google for access to its search index.
Google denied the request.
Bing’s results had “significant quality issues” and were “at best a short-term solution.”
Access to Google’s index would have sped up OpenAI’s development of its own search system.
Why It Matters
This revelation could fuel the rivalry between Google and OpenAI, two companies racing to dominate the future of AI-powered search. If proven true, it raises major questions about how AI companies source their data — and whether Google’s index is still the most valuable resource in the world of search.
