Select Language:
Earlier reports indicated that ChatGPT from OpenAI might be utilizing segments of Google search results to craft its responses—thanks to the digital marketing community for catching that early. Now, according to The Information, OpenAI has been partially sourcing current event data, such as news and sports updates, by scraping Google search results through a startup called SerpApi.
The article notes that SerpApi is an eight-year-old web scraping company, which publicly listed OpenAI as a client until as recently as May last year. The company then removed this reference, though the reasons remain unclear.
This approach isn’t limited to ChatGPT alone; another AI tool named Perplexity appears to be using the same methodology. A recent social media post confirms that SerpApi supplies data to both organizations, revealing that OpenAI relies on scraped Google search results for real-time information on trending topics.
Additionally, a recent study by Ahrefs reported that approximately 83.39% of ChatGPT’s top results are not found in Google’s search listings for the same queries. Despite this, there are often instances where responses from ChatGPT align with Google’s search outcomes.
Discussions about this practice are ongoing on various online forums, highlighting the emerging trend of AI models sourcing current data through search engine scraping techniques.
ChatGPT
Perplexity
Gemini AI
Grok AI
