Will we define global reporting standards for AI?

How to export sources and PDFs from your scienceOS library in 2 simple steps.

The use of large datasets, often containing sensitive personal information, presents significant privacy risks in AI development and application. Ensuring the confidentiality and security of this data is not just a technical requirement - it is a fundamental ethical responsibility.

Researchers rarely work within a single software environment. Literature discovered, analyzed, and organized in one tool often needs to be transferred into reference managers, writing environments, and publication workflows.

Avram Update: Your scienceOS library now supports more ways to import and organize your research.

You can upload any PDF with 128 pages or fewer and with a maximum file size of 80 MB on Free and Angel plans. On Team and Institution plans, the limit is 1024 pages and 80 MB.

Collection structures often evolve alongside a research project. As topics become more clearly defined and organizational systems mature, researchers may need to adjust collection names and visual identifiers to ensure their library remains intuitive and easy to navigate.

Science itself is complex and often a simple research design may not be suitable for a research question or field of science. However, this complexity must not apply to research tools. At scienceOS, we value effortless experiences, ensuring that researchers can navigate their work seamlessly.

When you upload a PDF to your scienceOS account, it is stored securely, encrypted automatically, and accessible only to you. If you create a project and add PDFs to it, you can choose to share access with colleagues and others.

ScienceOS partners with Comunidad BioLynx. Comunidad BioLynx is Spain’s leading community for young scientific talent, bridging the gap between university and academia and the pharmaceutical/biotech industry by equipping students with the skills and knowledge needed to succeed.

How to get a comprehensive overview with scienceOS in 2 simple steps.

We will be giving a talk about scienceOS at the de.KCD Industry Forum in Berlin on June 18, 09:00-17:30 (CEST). We will demonstrate how scienceOS enables secure exploration, comparison, and synthesis of scientific literature and proprietary data entirely within EU-hosted infrastructure.

Sci.STEPS published an interview with Henri Max Deda, CEO of scienceOS, about his path from molecular biology into building a company focused on work with scientific knowledge. In the conversation, he explains why connecting researchers with the data they need became such a compelling mission.

Accurate metadata is essential for working effectively with scientific literature. Titles, authors, and publication details are what allow researchers to quickly identify, interpret, and reference sources without repeatedly opening the underlying documents.

Does scienceOS train on my data?

How to verify the accuracy of citations in your manuscript with scienceOS in 3 simple steps.

Can we mitigate bias in AI responses? One key issue regarding AI is: bias and discrimination. If the data used to train AI models is biased - whether due to societal inequities or historical patterns in the case of social sciences - these systems can replicate or even amplify these biases.

How to generate an evidence table from dozens of papers with scienceOS in 3 simple steps.

Wiener Update: Deep research now adapts to your input in scienceOS.

In collaborative research settings, assembling relevant literature is rarely the responsibility of a single person. Sources are often distributed across individuals, institutions, and access points, making it difficult to bring everything together in one place.

AI ethics strategist Olya Vvedenskaya will be joining a panel discussion at the MCAA Annual Conference 2026 organized by Marie Curie Alumni Association on April 23, 10:30-11:30 (CEST).

Many complex problems require interdisciplinary approaches, vast amounts of resources, and a multitude of creative ideas. Therefore, collaboration is at the heart of research. At scienceOS, we believe in collaboration. We value each person's input and move forward together.

How to calibrate trust in AI effectively.

We are committed to strict EU data‑residency and processing controls, and provide a fully GDPR-compliant AI research agent. All our AI and machine learning models, as well as your chats, your uploaded files, and your user data are located on servers within the European Union.

A new paper on genAI in qualitative social sciences with scienceOS.

In scientific work, continuity of focus is essential. Researchers frequently move between reading, evaluating, and organizing sources, and even small interruptions can disrupt workflow and increase cognitive load.

With our invited speakers (Dr. Joss von Hadeln, Dr. Ulrich Degenhardt, Dr. Olya Vvedenskaya), we aim to tackle the risks of cognitive offloading, the high cost of verifying AI-generated content, the human tendency to trust conversational agents, and the definition of trust.

Sci.STEPS published an interview with Mark Reinke, co-founder and AI engineer at scienceOS, about his transition from jazz pianist and music teacher to computer scientist and AI engineer. In the conversation, Mark talks about his interest in building tools for retrieving and organizing knowledge.

How to tell if an AI tool is trustworthy.

We will be hosting a webinar about AI ethics in scientific research on April 10, 14:00 (CEST). Quick facts: Format: Free webinar with three short talks and a discussion round Date: April 10, 2026 at 14:00 (CEST) Sign up here: forms.gle/UnoxLBzDC1MT...