OpenAI has enhanced Deep Research within ChatGPT. This feature is now operational on the new GPT-5.2 model, as announced by OpenAI on X. A key addition is the ability for users to connect applications to ChatGPT and—potentially very useful—search specific websites. Search progress can be tracked in real-time, with the option to insert follow-up questions or add new sources. Results can now be displayed as full-screen reports.

Until now, Deep Research—launched in 2025—operated on the o3 and o4 mini models. OpenAI considers it the first "AI agent" in ChatGPT because the system independently initiates multi-stage web searches based on user queries before generating a response.

That said, even web searches cannot prevent generative AI from making mistakes; the longer the generated text, the higher the risk of errors. In daily use, targeted search queries using powerful reasoning models are generally more reliable. While web searches significantly reduce the incidence of hallucinations overall, they do not eliminate them completely.