ChallengeInstances

Application of AI has potential to exacerbate, or ignore, established injustices; these biases may not be immediately clear in unbalanced data, nor apparent in short-term implementation evaluations

"The research analytical process includes selecting the training data, cleaning the data, developing the model through steps of training, evaluating, adjusting, re-training the model.[..] The use of AI systems to uncover or predict social phenomena can be tainted by biases in the training data set on certain demographics or proxies thereof, which may lead to unfair and unjust outcomes.[...]. " (franzke, et al., 2020 p.41-44). Potential for biases and outcomes that do not respect justice arises throughout the research processes.

Challenges Research produces both direct impact, on participants and then society, while also giving rise to indirect, long-range, and dual-use impacts

Overarching Principles Justice Consider long-range and indirect impacts

Principles Justice Consider long-range and indirect impacts

Reference franzke, aline shakti, Bechmann, A., Zimmer, M., Ess, C., & Association of Internet Researchers (AoIR). (2020). Internet Research: Ethical Guidelines 3.0 Association of Internet Researchers. AoIR. https://aoir.org/reports/ethics3.pdf

Sources AoIR report 3

Strategies Questions to consider in key stages of AI and machine learning based research, regarding the Research Analytical Process

Created At 2023-06-16T12:12:34.000Z

Title Application of AI has potential to exacerbate, or ignore, established injustices; these biases may not be immediately clear in unbalanced data, nor apparent in short-term implementation evaluations