We are committed to providing safer access to Generative AI. Learn more about the principles that guide us The rise of Generative AI technologies has a continuing profound impact on society and the daily life of staff and students at The University of Edinburgh. We are committed to remaining at the heart of the world’s AI innovation, while maintaining a strong commitment to AI Safety, Responsibility and AI Ethics.The principles below outline our commitment to providing staff and students with safer access to Generative AI technology through ELM (Edinburgh (access to) Language Models).1. SafeWe are continually working on improving and building new safety features for ELM. ELM already has robust features that detect and remove harmful inputs and outputs.2. Secure We have a Zero Data Retention agreement with OpenAI and continue moving towards providing locally run Large Language Model access for even further enhanced data security. Using ELM is more secure than the popular alternatives or using ChatGPT directly.3. ResponsibleUsing Generative AI has an impact on the environment. We are moving towards providing an alternative to power hungry popular Generative AI models through access to locally hosted models. This will allow staff and students to lower their impact on the environment. It will also allow us to measure the impact more accurately.4. InclusiveWe are committed to ensuring generative AI is used for good across our University community. We are continually looking for new ways to assist with adoption and innovation to encourage inclusivity e.g. by provision of free API access.5. EquityHaving access to the newest and most powerful AI tools is expensive. We provide access to tooling at no cost for all staff and students. This is to ensure that we are levelling the playing field and allowing everyone to leverage these tools regardless of their economic situation.6. ResponsiveWe are continually working with our university community to ensure compliance with the university guidelines and to understand the unique needs of all stakeholders.Want to learn more about our AI Principles in practice? Click on our Responsible AI & AI Safety pages Return to ELM main page This article was published on 2025-01-17