The University of Michigan researchers chose X for their study on AI-generated NCII, citing the platform's lack of volunteer moderators and limited impact from paid moderators, which would allow them to explore content moderation processes without significant interference.
Researchers ensured AI-generated NCII did not re-traumatize victims by testing images against facial-recognition software and reverse-image lookup services, confirming a total lack of resemblance to real individuals, thereby emphasizing their commitment to ethical standards.
The study implies that X has failed to effectively moderate non-consensual nudity, as flagged content was not acted upon without DMCA intervention, further stressing the need for accountability in content moderation strategies.
The researchers believe their study could prompt social media platforms like X to enhance moderation practices for NCII, highlighting potential benefits that could outweigh the ethical concerns raised by their experiment.
Collection
[
|
...
]