At the AI Ethics Lab, we developed our Ethics in R&D (ERD) model to integrate ethics into AI research and development.
Our model consists of 3 components, which together provide a comprehensive approach to ethics. Yet, each one of these components can also be implemented in isolation. These are,
(1) Understanding Ethics: researchers’ and developers’ awareness and understanding of ethics
(2) Ethics Analysis: embedding ethical analysis into the design and development process
(3) Institutional Ethics Policy: developing institutional policies for recurrent crucial ethical questions
We argue that the ERD model provides a meaningful way of integrating ethics into innovation, where ethics complements and enhances R&D process. In our model, ethicists are not sitting in an isolated board – they are trusted collaborators of researchers and developers, working together to improve technologies.
ERD model can be implemented fully (with all 3 components) or partially within companies, start-ups, research centers, and incubators.
TRAINING IN ETHICS
As developers and researchers build and create AI systems, they are well positioned to catch ethical problems. In our ERD model, we recognize developers’ and researchers’ unique position to be effective “first responders” for ethical problems and we train them to recognize and flag them as they work through a project.
Through our trainings and workshops, we ensure that they understand
- the ethical concerns around the technologies that they are developing,
- the connections and conflicts between these ethical concerns, and
- how these ethical concerns feature in practical application.
Ethics involves analytic, structured, and systematic reasoning. We developed modules and workshops that enable non-ethicists to become familiar with ethical reasoning and learn to use this tool.
To request an assessment of your company / institution and discuss implementing ERD, please contact us at firstname.lastname@example.org.