Our Responsible AI (RAI) services present a comprehensive and integrated approach to AI ethics. When applied together, they cover all aspects of AI ethics within an organization when developing or deploying AI systems.
Our RAI services are based on the Puzzle-Solving in Ethics (PiE) Model developed by AI Ethics Lab, an independent collaborator of EAI. The PiE Model lays out four core components of AI ethics implementation: Roadmap, Strategy, Analysis, and Training. Using EAI’s resources and expert network, we expand the PiE Model with additional services: Advisory Board, Technical Audit, and Systems Registry. Each of our RAI services may be used as a stand-alone component or we can work with you to combine them in a mix that best suits your organization.
AI Ethics Roadmap
The AI Ethics Roadmap is the first step in implementing a Responsible AI practice. It provides an organization with insights on the most pressing needs and long-term goals for integrating and implementing ethics into its innovation cycle. To map out the Roadmap, we examine the structure and the existing ethics mechanisms within the organization. Based on this assessment, we design an action plan.
AI Ethics Strategy
The goal of the Responsible AI practice is to put in place mechanisms that upskill the organization for developing and deploying AI systems ethically. These mechanisms must be in line with its other stated values, principles, and organizational structure.
By customizing the AI Ethics Strategy, we ensure a seamless and efficient implementation of ethics in innovation. The Strategy consists of three P’s: (i) Playbook, (ii) Process, and (iii) People.
AI Ethics Analysis
Conducted by ethics experts, AI Ethics Analysis aims to determine the ethics risks related to the domain, the technology or system, and, at the intersection of both, to the specific product or project. The Analysis focuses on the ethical risks but also lays out the ethical opportunities that the innovation presents. Analyses in different phases of innovation will have varying degrees of focus and depth.
AI Ethics Training
Sustainable implementation of a Responsible AI practice is only possible with a network of skilled employees at all levels of the organization. We customize AI Ethics Training for an organization’s needs, domain, and goals.
Independent AI Ethics Advisory Board
EAI brings together world experts in AI ethics from diverse backgrounds and profiles to create a high-level advisory board.
Technical AI Audit
Technical audits are essential parts of the Analysis process used to determine a system’s ethical robustness. EAI’s highly qualified data scientists and computer engineers dive deep into the AI system, evaluating it for its ethically crucial aspects such as bias, appropriate use of data, security, and accuracy.
AI Systems Registry
Registering all the components of an AI system simplifies documentation and maintenance. Furthermore, it provides system transparency if done publicly and simplifies future algorithmic audits.