Chatbots and search engines may both aim to deliver accurate answers, but a new study shows that they operate in strikingly different ways. While Google’s search results are driven by ranking and visibility, AI chatbots appear to look far beyond the most popular pages on the internet.

Researchers from Ruhr University Bochum and the Max Planck Institute for Software Systems (via Tom’s Guide) examined how Google’s search engine compared with Google’s AI Overviews, Gemini 2.5, and GPT-4o’s web-based results. They tested a range of queries, from political questions and factual information to product recommendations, to see where each system sourced its answers.

Also read: Google raises red flag on AI scams fooling job seekers and small businesses

AI ChatBots vs Google: Different Paths to the Same Answer

The findings revealed that AI chatbots often collect data from a much broader range of online sources than traditional search engines. In many cases, they accessed web pages far beyond Google’s first 1,000 results, sometimes even from domains that ranked outside the top one million websites.

When it came to shopping-related searches, the overlap between Google’s top results and AI-generated ones was less than 30 percent. Across all query types, the similarity stayed below 50 percent. Gemini showed a particular tendency to draw content from low-ranking or lesser-known websites.

Also read: iPhone Air 2 tipped to come with dual camera setup with new upgrades- All details

While this might suggest that chatbots rely on less trusted material, the researchers found that this was not necessarily the case. GPT-based systems often cited formal or verified sources, including encyclopedias and corporate sites, and avoided social media references. These models also used online data to reinforce existing internal knowledge rather than start from scratch.

In contrast, Google’s search engine is built to assume no prior understanding from the user, ranking web pages by relevance, popularity, and optimisation.

Also read: iPhone 18 Pro tipped to follow Nothing’s transparent rear panel design

What It Means for Users

The study also noted that AI systems focus on depth and accuracy rather than presentation. They can extract insights directly from research papers or lengthy reports without simplifying the content. Their goal is to identify material that is reliable and informative, regardless of where it appears in search rankings.

The researchers did not declare one approach superior but emphasised the need for new evaluation methods to understand how generative search systems choose their sources. As AI continues to evolve, how these systems gather and filter information remains one of their most intriguing mysteries.