Microsoft’s New AI-Enhanced Bing Search Set to Challenge Google’s AI Overviews
Microsoft Launches AI-Driven Bing Generative Search, Posing a Challenge to Google’s AI Summaries.
Microsoft has unveiled a new AI feature for its Bing search engine called Bing generative search. This feature provides a concise snapshot of information in response to search queries, akin to Google’s AI Overviews introduced in the US in May. In addition to displaying relevant information, Bing’s new feature cites the sources from which the content is derived. The company is currently deploying this feature to a limited number of queries to minimize issues like AI hallucinations, which plagued Google’s AI Overviews.
Bing Generative Search Features
Microsoft announced the Bing generative search in a blog post, stating, “Today, we’re excited to share an early view of our new generative search experience which is currently shipping to a small percentage of user queries.” This feature builds upon the AI-powered chat answers launched on Bing in February 2023.
Although Microsoft did not disclose the specific AI models used, the feature leverages both large and small language models (SLMs and LLMs). The AI is designed to contextually understand search queries and process millions of sources to generate accurate and relevant search results.
The visual design of Bing generative search closely resembles Google’s AI Overviews, presenting information in a grey box at the top of the Bing search results page. Each snapshot includes a headline, a document index for easy navigation, and source citations. Additionally, the AI can incorporate images, videos, and tables when appropriate.
When Bing generative search results are displayed, traditional search results are moved to the right side of the page. The left side displays the snapshot first, followed by source links and related search results. Microsoft has emphasized its commitment to optimizing accuracy and is monitoring the feature’s impact on publisher traffic. However, the company has not revealed details about the AI model and architecture used for this feature.
