Physiopedia:AI Training Model Licensing Policy An Open Letter

As a UK registered charity, Physiopedia’s core mission is to advance education in physiotherapy and rehabilitation globally through the creation and dissemination of expert knowledge. We achieve this by building an open-source, collaboratively written online encyclopaedia covering a wide range of physiotherapy and rehabilitation topics. Over the years, our dedicated team has curated a comprehensive dataset of high-quality educational content. This dataset represents a significant intellectual property asset which is owned by the charity and its licensors.

In light of the growing interest in using such datasets to train and develop large language models and foundation AI systems, Physiopedia has developed a formal licensing policy. Our goal is to make this valuable educational resource available to AI companies and researchers in a manner that generates funds to support our charitable work, whilst preventing its unauthorised use without our permission or compensation.

We believe strongly in the importance of respecting intellectual property rights and copyright protections for creators and rights holders. The educational content in our dataset represents countless hours of work by our contributors who are subject matter experts in their fields. This intellectual property has significant value, which is why it is crucial that it is used only with proper authorisation, compensation and attribution. Physiopedia is committed to upholding intellectual property rights not just for our own purposes, but to promote a healthy ecosystem where creators’ rights are respected.

Physiopedia’s terms of use (which are published on our website) govern the use of our content and explicitly forbid, amongst other prohibited acts, web scraping, data mining, or any other unauthorised bulk extraction of our content for the purposes of training AI systems, or other commercial uses, without a valid license. Unauthorised use, copying and dissemination of our content in the training, development or commercialisation of AI systems deprives the charity of control over and appropriate compensation for the uses of our content, unreasonably prejudices our legitimate interests, and infringes our intellectual property and other rights. We reserve all rights in relation to any such unauthorised act.

Physiopedia is open to granting licenses to use our dataset for specific AI related purposes, such as training and fine-tuning, combining with other datasets, and commercialising specific AI products. We offer flexible licensing options and are open to receiving proposals. Organisations interested in licensing the Physiopedia dataset for AI training purposes should contact [email protected].

Our goal is to strike a balance – making our unique knowledge base available to accelerate AI development, while also upholding our charitable duty to generate funds that can be reinvested into creating more educational resources for the global rehabilitation community.

We welcome the opportunity to discuss potential partnerships that align with our charitable objectives. In this regard please note that we have appointed our affiliated company, Physiopedia International SL to conduct commercial negotiations and to conclude licensing arrangements on our behalf.