OpenAI is experimenting with a new search-focused prototype called SearchGPT that combines artificial intelligence with live information retrieval, marking the company's latest push into territory traditionally dominated by Google.
The prototype delivers answers to user queries with speed and freshness, pulling from current sources rather than relying solely on the training data embedded in its base model. Results appear alongside clear attribution to the original sources, addressing one of the persistent criticisms of AI chatbots: their tendency to produce plausible-sounding but unverified information.
The feature underscores OpenAI's ambitions to compete directly in search. Rather than simply generate text based on patterns learned during training, SearchGPT aims to provide users with immediate, sourced answers to questions that demand up-to-the-minute accuracy. This is particularly useful for queries about breaking news, stock prices, weather, and other time-sensitive topics where outdated information becomes quickly worthless.
SearchGPT remains in testing phase for now, described by OpenAI as a temporary prototype. The company has not announced broad rollout plans or indicated whether the feature will become a permanent part of its products. The emphasis on source attribution reflects growing pressure on AI companies to help users distinguish between information generated by algorithms and information drawn from real reporting and data.
The move comes as multiple tech companies race to integrate search capabilities into their AI systems. Microsoft has already embedded Bing search into its Copilot products, while Google has been layering AI summaries atop its traditional search results.
Author Emily Chen: "OpenAI's move into live search is smart, but sourcing matters far more than speed when you're competing against Google."
Comments