Page 153 - AI Computer 10

Value-based Frameworks

Value-based frameworks provide necessary ethical principles and values that guide decision-making across
various contexts, including the development and deployment of artificial intelligence (AI). These frameworks
prioritize fundamental values such as respect, fairness, accountability, and transparency, creating a moral
foundation for actions and decisions.
Value-based frameworks for AI encourage individuals and organizations to critically reflect on the moral
implications of their AI-related decisions.
The value-based frameworks are further classified into three main categories:
u Rights-based: A Rights-based framework prioritise the safeguarding of human rights and dignity, placing
a higher value on human life than any other concern. It underscores the necessity of honouring individual
autonomy, dignity, and freedoms. In the context of AI, this could mean ensuring that AI systems do not
infringe upon human rights or exhibit discrimination against specific groups.

u Utility based: A Utility-based framework helps to assess actions according to the principle of maximising
utility or the overall good, striving to achieve outcomes that provide the greatest benefit while minimising
harm. This framework aims to maximise overall utility or benefit for the largest number of individuals. In
the context of AI, this framework helps to evaluate the potential advantages of AI applications against the
societal risks they may pose, such as job displacement or privacy issues.

u Virtue-based: A Virtue-based framework focuses on the character and moral virtues of individuals involved
in decision-making processes. It examines whether the actions of individuals or organisations are consistent
with virtues such as honesty, compassion, and integrity. In the context of AI, virtue ethics could entail
evaluating whether developers, users, and regulators maintain ethical values throughout the entire AI
project cycle.

19
19