Don’t miss Prism as one AI firm has shown the importance of being earnest when it comes to research.
OpenAI has launched Prism, an AI workspace powered by GPT-5.2 that integrates directly into scientific research workflows.
This is a cloud-based environment that offers unlimited projects and collaborators, and is accessible to anyone with a ChatGPT account.
The platform represents OpenAI’s objective to dominate the scientific research field, with the chance that every future paper will essentially have ChatGPT as a co-author. Make of that what you will.
Kevin Weil, OpenAI’s VP for Science, said on LinkedIn, “2026 will be for AI in science what 2025 was for AI in software engineering.”
A ChatGPT moment
Prism embeds GPT-5.2 directly into the research environment where it maintains awareness of entire project contexts including methodology, data sources, and theoretical frameworks. The system can handle everything from initial abstracts through publication preparation, converting whiteboard diagrams into professional TikZ diagrams.
ChatGPT already receives approximately 8.4 million weekly messages related to advanced hard sciences, while breakthroughs from December showed AI solving Erdős problems and establishing proofs for statistical axioms using GPT-5.2 Pro.
A game of Crixet
Prism builds on Crixet, a cloud-based LaTeX platform that OpenAI acquired and transformed into this unified research tool. The workspace offers capabilities including claim assessment, prose revision, research discovery, and context management. Researchers can sketch on digital whiteboards and have the model assemble diagrams for figures, while the system manages references and summarizes all research related to papers.
Perhaps there will be early adoption in mathematics, theoretical physics, econometrics, and computational biology, with typical tasks including sanity-checking derivations, clarifying notation, and translating whiteboard work into publication-ready figures. Prism supports unlimited collaborators without seat limits or access barriers, allowing research teams to work together. More powerful AI features will be made available through paid ChatGPT plans over time.
What this means for scientific discovery
This launch signals a fundamental shift in how scientific research operates, with three key indicators that will determine Prism’s impact: measurable cuts in time-to-submission, lower revision cycles, and integrations with existing tools. However, significant challenges remain, including potential hallucination of citations, oversimplification of edge cases, and concerns about privacy of unpublished manuscripts.
What’s becoming clear is that Prism arrives as academia wrestles with both opportunity and risk, and tools that standardize reasoning steps, enforce citation checks, and keep context anchored could help reduce sloppy errors.
OpenAI has invested heavily in demonstrating scientific use cases, releasing papers on mathematical discovery, cell analysis, and biology experiments. The company aims to launch a “research intern”-level tool by September and automated AI research by March 2028.
Here are the top eight things you should never paste into an AI chatbot, ranked from easy-to-overlook mistakes to the most dangerous exposures.