OpenAI upgrades Deep Research to allow directing ChatGPT’s research toward specific websites and apps, with real-time progress tracking and the ability to adjust sources during report generation.

New output formats now include full-screen reports, tables, charts, and visual summaries to meet rigorous academic and professional needs.

Deep Research now runs on GPT-5.2, consolidating information from multiple sources into a single, accessible document.

From a business perspective, the polish and enterprise-ready outputs strengthen the value proposition for Pro and Enterprise subscribers by reducing post-processing and speeding workflow integration.

Analysts highlight economic and policy implications, including potential GDP impact, tiered pricing, upskilling needs, and calls for global standards to curb agentic AI biases and ensure accountability.

A Skills layer could let teams standardize workflows and share procedures, delivering consistent results without a full custom agent stack, though timing remains unclear.

The upgrade aims to reduce AI hallucinations in long outputs, while acknowledging that errors can still occur.

OpenAI’s move signals a shift toward outputs that are ready for professional workflows, potentially competing with traditional research services and consulting tools.

Deep Research has shown improved benchmarking performance and credibility for professionals, including legal practitioners, though occasional factual errors are noted.

Competitors like Google Gemini, Perplexity AI, and Anthropic Claude are enhancing structured, citation-rich outputs, prompting OpenAI to boost usability and workflow integration.

Targeted queries with capable reasoning models tend to be more reliable in practice.

Disclosure notes that Ziff Davis, PCMag’s parent company, filed a lawsuit against OpenAI in April 2025 over alleged copyright infringement related to AI training and operation.