Fairness-related harms
There are a number of types of harm that can be caused by AI systems when their creators have failed to take fairness into account. Fairness-related harms refer to the various types of negative impacts that result from AI systems when fairness considerations are not properly addressed during their design and development. These can include unequal distribution of benefits and drawbacks, unequal QoS, and perpetuation of harmful stereotypes and biases. Also, AI harms overlap because systems often cause multiple forms of harm simultaneously.
Let’s now discuss some key negative consequences that may occur when fairness is not considered during the design and development of AI systems:
- Allocation harms refer to the negative consequences that occur when AI systems provide or restrict access to opportunities, resources, or information. This leads to unequal treatment and prejudice against certain demographic groups, negatively impacting their ability...
