OpenAI Upgrades ChatGPT Research Tool
OpenAI has upgraded its Deep Research tool within ChatGPT, adding granular source controls and a full-screen viewer. The update allows users to focus AI-driven research on specific, trusted websites. This enhancement is a step toward providing more explainable and auditable AI for professional use cases like compensation analysis and HR compliance.
- The Deep Research feature, first introduced in February 2025, is powered by a specialized version of OpenAI's o3 model, which is designed for more complex reasoning tasks than standard models. A "lightweight" version for less intensive queries runs on the o4-mini model. - While the tool can take between five and 30 minutes to generate a report, the recent update allows users to monitor progress in real-time and make edits mid-run, such as adding new sources or adjusting the scope. - Usage is metered across different subscription tiers; as of June 2025, ChatGPT Pro subscribers ($200/month) receive 250 queries per month, while Plus, Team, and Enterprise users get 25. Free users are allotted five "lightweight" queries per month. - Before starting a task, the tool presents a proposed research plan which the user can review and modify, offering a layer of human oversight on the AI's intended actions and sources. - The launch of Deep Research is seen as a competitive response to other AI models and was announced shortly after the release of Chinese AI chatbot DeepSeek, which briefly surpassed ChatGPT in app store downloads. - This tool is part of OpenAI's broader strategy to deploy more autonomous "AI agents" for enterprise use, fitting alongside its 'Frontier' platform, which is designed to manage fleets of AI agents that can act as "AI coworkers" integrated with business systems. - For HR and compensation professionals, the ability to restrict research to trusted data sources like specific regulatory or salary survey websites is critical for tasks like benchmarking compensation and ensuring compliance with pay transparency laws.
