International Center for Ethics in the Sciences and Humanities (IZEW)

VIKING

Trustworthy Artificial Intelligence for Police Applications

Subproject:

Conditions for operationalizing ethical requirements for trustworthy AI in police applications.

The use of software solutions based on artificial intelligence is associated with a wide range of ethical issues. To be differentiated here is, on the one hand, the need for an ethical reflection on the definitions and the relationship of trust, traceability, transparency and fairness in the context of AI technologies. On the other hand, there is a focus on the operationalizability of ethical requirements with regard to the concrete application context. In the subproject "Conditions for Operationalizing Ethical Requirements for Trustworthy AI", a risk matrix as well as a catalog for ethical requirements for AI procedures will be developed.

Keywords: trustworthy artificial intelligence, traceability, attack resistance, de-biasing, transparency, ethical evaluation, standardization.

Researchers