OpenAI Launches Deep Research Mode to Revolutionize AI-Powered Internet Research

OpenAI has introduced a groundbreaking feature for ChatGPT called Deep Research, designed to perform complex, multi-step internet research. Announced during a livestream, this advancement pushes AI closer to agentic AI, where models autonomously handle human-like tasks.

A key highlight of Deep Research is its ability to access real-time internet data without latency constraints, though responses may take 5 to 30 minutes depending on task complexity. OpenAI stated that this feature enhances AI’s autonomy, marking progress toward artificial general intelligence (AGI)—AI that can perform intellectual tasks at a human level.

The new mode is tailored for knowledge-intensive work across industries like engineering, finance, literature, and biology. Initially, it will be available for OpenAI’s $200-a-month Pro users, with plans to roll it out to $20-a-month Plus users if safety standards are met within a month.

Deep Research is expected to deliver clear citations and transparent reasoning, reducing AI-generated misinformation. However, OpenAI acknowledged that hallucinations—incorrect or misleading responses—may still occur, albeit at a lower rate than previous models.

The launch comes shortly after OpenAI introduced o3 mini, a leaner reasoning AI model. Industry speculation suggests the rapid release of these models was influenced by China’s DeepSeek R1, a competitive AI model built with lower hardware requirements and a smaller budget.

OpenAI’s latest innovations highlight the accelerating race in AI development, with Deep Research poised to revolutionize online research and enhance AI’s capabilities for in-depth knowledge work.