Elsevier is expanding its LeapSpace research AI workspace with a broader foundation of licensed, peer-reviewed scientific content, while adding independent oversight for how its algorithms discover, rank and present research. The company says LeapSpace now combines Scopus records with full-text content from major publishers and societies, positioning the platform as a research-focused alternative to general-purpose generative AI tools.
Elsevier is expanding LeapSpace, its research-grade AI workspace, with substantially more full-text scientific literature and a new independent advisory board designed to provide oversight of the platform’s algorithms and content standards.
The expansion gives LeapSpace access to full-text research that Elsevier says represents 56% of global research citations and 42% of research articles published since 2021. The figures are based on Elsevier’s own analysis of Scopus data from June 2026, so they describe the company’s methodology rather than an independently audited market share.
LeapSpace already draws on 2.6 billion cited references from Scopus abstracts. The latest expansion adds full-text content through licensing agreements with publishers and scholarly societies including BMJ Group, JAMA Network, Rockefeller University Press and the American Society of Civil Engineers. Elsevier also expanded selected full-text agreements with existing partners such as Emerald Publishing, IOP Publishing and Sage Publishing.
The resulting content base combines Elsevier’s full-text peer-reviewed articles and books with more than 100 million scientific records in Scopus covering material from more than 7,000 publishers, according to the company.
That content strategy addresses one of the central challenges facing generative AI for scientific research: the difference between producing a plausible answer and producing an answer grounded in current, authoritative literature.
Elsevier’s own 2025 Researcher of the Future research illustrates the trust problem. Among more than 3,200 researchers surveyed across 113 countries, 84% said they had used AI, while only 22% considered AI tools trustworthy. Researchers identified transparent citations, current literature and high-quality peer-reviewed content among the factors that could increase confidence in AI research tools.
LeapSpace is designed around that research workflow rather than simply conversational question answering. Elsevier says researchers use the platform for tasks including designing study protocols, generating research data, writing manuscripts and theses, and preparing research for submission. In the company’s 2026 user research, nine in 10 surveyed LeapSpace users said the service was either essential to literature reviews or substantially contributed to them.
The platform’s positioning reflects a broader shift in enterprise and professional AI applications toward domain-specific systems. Gartner forecasts worldwide spending on AI models and platforms will reach $64.3 billion in 2026, up 63.4% from 2025, with spending increasingly focused on providers that can demonstrate performance, reliability, usage efficiency and measurable value.
For research AI, however, the relevant infrastructure extends beyond the underlying large language model. The quality, provenance, recency and licensing of the information supplied to the model can directly affect the usefulness of generated outputs. That puts AI data infrastructure, retrieval systems, scholarly search, citation management and research governance alongside model capabilities as competitive factors.
Elsevier is also attempting to address governance through the new LeapSpace Advisory Board. The independent panel includes computer science, AI and academic publishing experts from institutions including Carleton University, the University of Wisconsin-Madison, RWTH Aachen University, Leiden University, Zhejiang University and Japan’s National Institute of Informatics.
The board is intended to advise on algorithms used to surface and evaluate content, publisher-neutral search and ranking, responsible AI use in research, and transparent and explainable outputs.
That governance layer is notable because AI research tools face a different set of expectations from consumer chatbots. Researchers need to be able to trace claims to scientific literature, assess the relevance and currency of evidence, and understand where automated systems have influenced the research process.
Gartner’s research points to a similar challenge as AI systems become more autonomous. The analyst firm says only 13% of organizations believe they have the right AI-agent governance in place, while it expects an average Fortune 500 enterprise to have more than 150,000 AI agents in use by 2028.
Elsevier’s approach is narrower, but the underlying issue is comparable: specialized AI systems need governance mechanisms that match the environments in which they operate.
The company is also using LeapSpace as a workflow layer rather than limiting AI to literature discovery. Its stated goal is to support researchers across activities from literature reviews through manuscript preparation while keeping researchers involved in critical evaluation.
LeapSpace is available globally to institutions and corporations, with individual plans and free access for academics and students. Elsevier also says the platform won Best Generative AI Solution at the 2026 CODiE Awards.
The larger competition in AI for scientific research is likely to be shaped not only by model quality but by access to trusted data, licensing arrangements, citations, transparency and integration with established research workflows. As researchers move from experimenting with general-purpose AI toward using AI throughout the research lifecycle, those supporting layers could become as important as the model generating the response.
Market Landscape
The research AI market is moving toward domain-specific platforms built around authoritative datasets, retrieval systems, workflow automation and governance rather than generic chat interfaces alone.
Elsevier’s LeapSpace competes indirectly with research-oriented AI products from technology companies and specialist providers, while its access to licensed scholarly content gives it a different foundation from tools that primarily retrieve information from the open web.
The opportunity is significant. Gartner projects 2026 spending on AI models and platforms at $64.3 billion, up 63.4% year over year. At the same time, researcher trust remains a constraint: Elsevier’s survey found only 22% of researchers consider AI tools trustworthy.
Top Insights
- Elsevier is expanding LeapSpace with licensed full-text scientific content from major publishers and scholarly societies.
- The platform combines Scopus records with peer-reviewed articles and books to support research workflows beyond literature search.
- A new independent advisory board will provide guidance on algorithms, transparency, publisher neutrality and responsible AI use.
- Researcher trust remains a major challenge, with Elsevier reporting only 22% currently consider AI tools trustworthy.
- Domain-specific AI platforms are increasingly competing on data quality, provenance, governance and workflow integration alongside model performance.
Power Tomorrow’s Intelligence — Build It with TechEdgeAI
