Elsevier unveils LeapSpace to boost trusted AI in research
Elsevier has introduced LeapSpace, an AI-assisted workspace designed to streamline academic and corporate research, drawing solely from peer-reviewed scientific sources and integrating features to foster collaboration and enhance research outcomes.
AI trust challenges
According to recent findings, the majority of researchers are already using AI tools in their work, yet there remains a considerable gap in trust. While 84% of researchers report engaging with AI, only 22% say they trust existing solutions.
LeapSpace aims to address these concerns by underpinning all AI-generated insights with transparent sourcing and evidence strength indicators.
Trusted content sources
The platform brings together millions of peer-reviewed articles, book chapters, and research abstracts, including data from Elsevier's Scopus-its publisher-neutral database covering more than 100 million records from over 7,000 publishers worldwide.
Through this integration, LeapSpace offers researchers a single environment to generate ideas, plan projects, search literature, identify collaborators, and discover funding opportunities.
Transparency focus
One notable feature of LeapSpace is the inclusion of Trust Cards, which accompany every AI-generated answer.
These cards provide source references, explanations of citations, highlight contradictions, and allow users to calibrate the strength of evidence supporting each insight. An independent advisory board is responsible for ensuring the platform's transparency, with oversight of algorithmic explainability and publisher neutrality.
Workflow enhancement
For efficiency, researchers can rapidly evaluate evidence using tools like Reading Assistant and Author Search while also having the ability to upload their own material for deeper analysis. The funding discovery tool taps into 45,000 active and recurring grants, estimated at a combined value of over USD $100 billion. Elsevier states that the workspace is developed with enterprise-grade security and strict adherence to responsible AI and privacy standards.
Technical approach
LeapSpace employs a combination of agentic AI, generative AI, reasoning engines, and retrieval-augmented generation technology. These tools are intended to support a range of workflows including literature analysis, hypothesis generation, and identification of data gaps.
Every step prioritises transparency and human oversight, giving users clear documentation on how AI output is generated.
Community development
Thousands of researchers from more than 300 institutions across 64 countries participated in the development and testing of LeapSpace. Early feedback indicates that the platform has potential to reduce the time spent on research tasks, uncover previously missed insights, and improve research design.
"Rare disease research is limited by the small number of experts focused on each disease, Sanfilippo syndrome being no exception. We rely on connections and learnings across multiple disciplines to bridge gaps in direct evidence. Synthesizing vast amounts of often disparate information is a challenge. In my early experience with LeapSpace, I've been impressed with how it's able to address these challenges while providing confidence in the accuracy and rigor of its outputs," said Cara O'Neill, Chief Science Officer, Cure Sanfilippo Foundation.
