The context
This story is developed in the context of IEEE P7014 working group, which is involved in the development of a standard for the ethics of Empathic AI (Emotion AI, Affective Computing, etc.).
There has never been a greater need for ethical innovation, due to the risks associated with the advancing intelligence of new technology. The global community creating technology standards has always supported innovation while respecting planet and people. Now is the time to join this community and help the standards catch up with the tech.
Ben Bland, Chair, IEEE P7014.
The challenges
Various efforts to standardise autonomous and intelligent systems are being conducted in relative isolation, leading to a common desire for more integration and common resources between efforts. At the same time, a major ethical challenge for these advanced systems is how to make them explainable and interpretable for all stakeholders. IEEE P7014 is developing a common framework for AI explainability to bridge these gaps.
How standardisation activities help face the challenges
Autonomous and intelligent systems entail various potential ethical risks that could affect people or planet. With access to a central framework for explainability, our aim is that all stakeholders of these systems will be empowered to understand and interrogate them, and ultimately to benefit from a greater level of safety, privacy and agency.
The Benefits
The main benefits include:
- Systems developers can follow the framework to provide reasonable levels of explainability to their audience.
– Those working in governance, policy, review and certification roles should be able to assess the developer's explainability efforts.
– Users and the public should be able to expect adequate levels of explainability by those developers who follow the xplAInr framework.
– Developers of standards, laws and regulations should be able to apply the framework to make their work practicable for developers, consistent across regulation projects, and comprehensive in their coverage.
Future plans
xplAInr is still in an early version. After primarily focusing on the design of the framework on systems developers, the focus has been on expanding it to include governance and policy actors, as well as developers of standards. The main step for the future is to expand the utility of the framework to encompass users and the public, and continue to enrich it with practical resources such as reference standards, and best practice processes