Ethics of Cognitive Computing and Social Sensing

Context & Background

As cognitive computing systems behave and act like humans, they introduce critical ethical questions. AI models can inherit biases, profiling users on social media, violating privacy, and raising concerns about trust and accountability. It is essential to integrate ethical guidelines directly into the design thinking stage of AI systems.

Problems to be Addressed

Existing AI frameworks lack methods to compute or enforce ethical compliance. Algorithms are often 'black boxes' that cannot explain the moral reasoning behind their recommendations, especially in sensitive domains like healthcare and legal triage.

Aims and Objectives

1. Analyze ethical and societal challenges in cognitive systems.
2. Design algorithms using Inductive Logic Programming to evaluate ethical choices.
3. Produce policy guidelines and frameworks for industry adoption.

Methodology

The project combines philosophical ethics (deontology, virtue ethics, justice) with machine learning. The team develops an ontology of ethical concepts and builds a model that learns from positive and negative examples of moral choices. Inductive logic programming is chosen to ensure the ethical decision-making process is fully explainable.

Expected Outcomes

A software toolkit for ethical scoring of AI, policy white papers, workshops with industry leaders, and research papers on AI governance.