OpenAI rolled out two new web‑enabled modes for ChatGPT, aimed at turning the chatbot into a faster, more reliable research assistant. The “search” option pulls the latest web results into a conversation, letting users ask for any immediate current events, market trends or niche facts without leaving the chat window. A globe icon flags responses that rely on live data.

Deep research goes a step further by orchestrating a multi‑step inquiry across the internet, evaluating sources, refining queries and synthesizing a thoroughly structured report with citations. Users launch it from the tools menu, supply a detailed prompt and receive a ready document after 5‑30 minutes, and complete with actionable insights and risk assessments. Enterprise admins can toggle the feature on or off.

The two modes complement each other: search delivers quick, up‑to‑the‑minute facts, while deep research tackles complex, multi‑source questions that would otherwise require manual digging. Both embed source links, urging users to verify information before acting. By blending real‑time web access with the model’s reasoning, OpenAI positions ChatGPT as a hybrid analyst for cross‑functional internal large business, academic and product teams.