Help Centre

Springer Nature is committed to delivering responsible AI enabled solutions to support the global research community. Feedback from our researchers is clear: they look to us not only for guidance and support in using AI. This is why we have developed Nature Research Assistant.  

‍

Nature Research Assistant responds directly to what researchers have asked us for – a trusted AI-powered tool that supports them in doing their work, but better. Built with - and specifically for – researchers, Nature Research Assistant uses trusted content and expertise to help researchers develop, improve and refine their work in line with the highest editorial standards. It currently supports researchers undertaking everyday tasks such as writing, reading and discovering research papers.

‍

Nature Research Assistant does not store or use any content uploaded by researchers to train any other AI tool or model.   

‍

Nature Research Assistant has already been extensively tested by researchers, but it is now in beta testing with additional researchers to further test its ‘real-world’ application.  Feedback from this phase will continue to inform, increase and improve its functionalities. We are actively encouraging those testing it to provide feedback on Nature Research Assistant and its outputs to help us improve the tool for researchers around the globe.

Nature Research Assistant is designed to help you discover, read and understand research papers faster. You can also use it to get advice on writing original research manuscripts.      

You should only use Nature Research Assistant for research purposes to assist you in identifying, understanding, and crediting source material relevant to your research.  Nature Research Assistant's outputs are not a substitute for your own research.  You should refer to the underlying source material for the full context and continue to follow usual standards of academic integrity while using Nature Research Assistant.

Nature Research Assistant is powered by AI. Outputs are generated by large language models (LLMs) and we use a range of in-house developed machine learning techniques to generate the high-quality outputs. The content users upload is not used to train the LLM. The features have been carefully designed by domain experts to meet the specific needs of researchers and to ensure high quality outputs.   

‍

AI can make mistakes – Nature Research Assistant may occasionally generate incorrect or misleading information. When using the tool, researchers should always refer to the underlying source material for the full context and double check outputs, including against the underlying source material, before using, sharing or publishing them. 

Nature Research Assistant is not a generic AI tool—it’s built specifically for researchers. A recent Springer Nature survey found that over 50% of researchers see AI as increasingly important in supporting their work. However, they want tools they can trust—tools that protect their data, uphold scientific integrity, and are developed with their needs in mind. Data also shows that researchers are looking to publishers to help support and guide this. 

‍

Nature Research Assistant has been developed in direct response to these needs, and in collaboration with researchers. It’s powered by AI and built on a vast corpus of peer-reviewed academic content. Its features are designed to save researchers time without compromising scientific rigor. 

‍

Importantly, user content is never used to train the AI, and we maintain strict standards around data use and AI integrity. While AI outputs should always be reviewed critically, and it is always important for researchers to read the full paper of any that they cite, the team at Nature Research Assistant have developed measures to enable outputs to be of high quality based on a range of criteria.  

‍

For example, what sets Nature Research Assistant apart from ChatGPT? 

‍

No, Nature Research Assistant is publisher-agnostic and can be used with any paper.

Although the tool aims to improve the quality of your submission, you should refer to the guidelines for the journal or publisher you wish to publish your research with to check their latest policies on the use of AI tools.

‍

You can find the latest policy for journals published by Springer Nature here.

‍

As of May 2025, Nature Portfolio’s guidance states: “Large Language Models (LLMs), such as ChatGPT, do not currently satisfy our authorship criteria. Notably an attribution of authorship carries with it accountability for the work, which cannot be effectively applied to LLMs.   Use of an LLM should be properly documented in the Methods section (and if a Methods section is not available, in a suitable alternative part) of the manuscript. The use of an LLM (or other AI-tool) for “AI assisted copy editing” purposes does not need to be declared. In this context, we define the term "AI assisted copy editing" as AI-assisted improvements to human-generated texts for readability and style, and to ensure that the texts are free of errors in grammar, spelling, punctuation and tone. These AI-assisted improvements may include wording and formatting changes to the texts, but do not include generative editorial work and autonomous content creation. In all cases, there must be human accountability for the final version of the text and agreement from the authors that the edits reflect their original work.” Please check for the latest guidance.  

Please note that using Nature Research Assistant in no way implies that any article written by you will be selected for peer review or accepted by a Springer Nature publication or any other publication. We do not and will not intervene in manuscript selection by journal editors. The editorial decisions of any journal editor are discretionary and based on the quality and suitability of a manuscript for that journal and as such are entirely independent of this tool.  

‍

You can find more information on recommendations for AI use in manuscripts here.

Nature Research Assistant has already been extensively tested by researchers. Additional researchers can now join the waiting list for access. Feedback from this phase will continue to inform, increase and improve its functionalities. We are grateful to the researchers who are using the tool for their feedback which helps us to improve Nature Research Assistant for researchers around the globe.

You can upload your own papers, and other papers that you have the necessary rights to use.

‍

Before uploading any content, please check and ensure that you have the rights to:

• upload the content to Nature Research Assistant.

• use the content in the manner you anticipate and for the uses you ask Nature Research Assistant to carry out. 

• In order to ensure you have the necessary rights, you may need to review your terms of access to the content and any arrangements you have with your institution or employer.   

If you are not sure you have the necessary rights to use any particular content with Nature Research Assistant, please do not upload it.   

‍

For the avoidance of doubt, if your institution or organisation subscribes to Springer Nature content, you are able to use this content in the tool.

The content that you upload will not be available to the Publisher (Springer Nature) or others. The content is not used to train the Large Language Model (LLM) and any papers that you upload, including draft manuscripts, will not be shared with any active editors or other researchers.

Nature Research Assistant draws from 23 million papers from a combination of Springer Nature journals content, Elsevier content and openly available content from a wide range of credible academic publishers from 2000 to 2025.

Springer Nature stores published content in our vector base, and we use semantic search and other machine learning techniques to search for relevant papers. The LLM is not trained by the content used in the search.

‍

Having your content in the vector base means that it will be discoverable to other researchers. For example, it may be recommended as further reading, meaning more researchers can discover your work. Users must have the rights to view the full text of your article. Nature Research Assistant does not allow users to circumvent paywalls.

You can use our prompt library within the tool, which contains the prompts we consider produce the best results from Nature Research Assistant.  You can also write your own prompt. Here are some tips to help you with prompt writing:  

• Consider how you craft your prompt in terms of bias, both in terms of input and output.

• Do not include personal, sensitive, or confidential information in prompts.  

• Do not create prompts that could lead to harmful or malicious outcomes.  

• Take accountability for the impact of your prompts.  

• Before deploying prompts widely, test them to ensure they lead to appropriate and ethical responses. Be guided by our AI Principles.

• Stay informed about the latest developments in AI ethics and incorporate best practices into prompt creation.  

We are committed to making Nature Research Assistant as accessible as possible to everyone, including those with visual, hearing, cognitive and motor disabilities. We're constantly working towards improving the accessibility of our website to ensure we provide equal access to all of our users. For further information please see our accessibility statement.

Nature Research Assistant is currently in a Beta testing phase. We are continuing to develop Nature Research Assistant to increase and improve its functionalities. If there is anything that we can do better, we encourage you to reach out and provide us feedback by contacting us at researchassistant@nature.com.  

Please contact us at researchassistant@nature.com.