Statement on the use of Artificial Intelligence (AI)

Artificial intelligence is reshaping how science is done, how evidence is produced and shared, and how societies make decisions. As the independent scientific academy of the United Kingdom, the Royal Society will engage with these technologies in a way that advances knowledge and public benefit while upholding the highest standards of integrity and care. We will proactively use AI in our work according to our values:

• We make a difference. We will harness AI where it can strengthen research quality, accelerate discovery, improve access to scientific knowledge, help address major societal challenges, and accelerate our support to scientists. Our priority is to ensure that AI is applied in ways that produce clear public benefit and that scientific judgements remain grounded in rigorous evidence and peer scrutiny. 
• We are collaborative. AI work at the Royal Society will be driven by collaboration — across disciplines, with our Fellowship, with policymakers, and with public and private partners. We will share learning, tools, and standards where doing so advances trustworthy and reproducible science. Collaboration also means listening to diverse perspectives about the risks, limitations, and ethical implications of AI. 
• We strive for excellence. All outputs that rely on AI will meet the same standards of excellence that define our work: accuracy, transparency, reproducibility, and methodological rigour. We will require appropriate human oversight and specialist review for AI-assisted findings while being open to the areas where AI can improve our ability to process high volumes of information. We will be clear about the provenance and limits of AI contributions. 
• We respect each other. The Royal Society is committed to fairness, inclusion, and respect in how we develop and use AI. We will guard against misuse of personal or sensitive data, ensure that AI is not used to entrench bias or exclusion, and protect the privacy and dignity of individuals – including intellectual property and copyright - involved in our work. Our approach will be consistent with our existing codes of conduct and ethical standards. 

We want to encourage staff to use AI in a way that supports their work and provides them with innovative ways to deliver, enhancing the Royal Society’s reputation and impact. AI use comes with a responsibility to protect our values, and we will adopt the following guiding principles in support of this: 

• Open-mindedness. We will proactively identify the areas where AI can improve our efficiency and effectiveness, especially to free staff from onerous tasks.
• Human accountability. People, not systems, remain responsible for our most critical scientific judgement and decisions.
• Transparency. We will be open about where and how AI is used and will document methods and limitations.
• Data stewardship. We will use and store data ethically and securely, consistent with legal obligations and best practice.
• Quality assurance. AI outputs must be validated, reviewed and reproducible.
• Adaptive governance. We will maintain oversight, provide staff guidance, and update our approach as technologies and norms evolve.

This statement is the Royal Society’s public commitment to principled AI use. It supports and extends our existing positions on AI, such as our policy towards the use of generative AI in funding applications and assessment. (See Grants policies and positions). We will publish a detailed internal policy and practical guidance for staff to translate these principles into everyday practice and to support colleagues in responsible adoption.