New OpenAI tool renews fears that “AI slop” will overwhelm scientific research
On Tuesday, OpenAI released a free AI-powered workspace for scientists. It’s called Prism and immediately sparked skepticism among researchers who fear the tool will accelerate the already overwhelming flood of low-quality articles in scientific journals. The launch coincides with growing concern among publishers about what many call “AI slop” in academic publishing.
To be clear, Prism is a writing and formatting tool, not a system for conducting research itself, although the broader OpenAI narrative blurs that line.
Prism integrates OpenAI’s GPT-5.2 model into a text editor based on LaTeX (a standard used for document composition), allowing researchers to write articles, generate citations, create diagrams from whiteboard sketches, and collaborate with co-authors in real time. The tool is free for anyone with a ChatGPT account.
“I think 2026 will be for AI and science what 2025 was for AI in software engineering,” Kevin Weil, vice president of OpenAI for Science, told reporters at a press briefing attended by the MIT Technology Review. He said ChatGPT receives about 8.4 million messages per week on “hard science” topics, which he described as proof that AI is moving from curiosity to the primary workflow for scientists.
OpenAI built Prism on technology from Crixet, a cloud-based LaTeX platform that the company acquired in late 2025. The company envisions Prism helping researchers spend less time on tedious formatting tasks and more time on actual science. In a demonstration, an OpenAI employee showed how the software could automatically search and integrate relevant scientific literature and then format the bibliography.
But AI models are tools, and any tool can be misused. The risk here is specific: by making it easier to produce polished, professional-looking manuscripts, tools like Prism could flood the peer review system with articles that do not significantly advance their field. Barriers to producing scholarly texts are decreasing, but the ability to evaluate this research has not kept pace.
