The discord between Stack Overflow users and the platform’s collaboration with OpenAI has reached a boiling point, with users expressing a strong preference to withdraw their contributions rather than see them utilized to train ChatGPT.
This outcry follows the recent revelation that OpenAI would harness the collective wisdom accumulated over 15 years from Stack Overflow’s user base to refine AI models. While Stack Overflow gains access to OpenAI’s advanced technology, OpenAI benefits from a treasure trove of data to enrich its neural networks.
Among the dissenters is Ben Ui, a software developer, who chose to voice his opposition by editing his previous contributions to protest the partnership. However, his efforts were met with a week-long suspension and a restoration of his posts to their original state, according to Ui.
“Stack Overflow prohibits the deletion of questions with accepted answers and significant upvotes to preserve community knowledge. Hence, I altered my top-rated responses to convey my dissent,” Ui explained on Mastodon. “Yet, moderators swiftly reverted my changes and suspended my account for seven days.”
Similar grievances have been voiced by others, alleging Stack Overflow’s wholesale bans on users who attempt to modify or erase their own content. Despite inquiries seeking clarification on the extent of these protests, Stack Overflow has remained silent.
The platform’s terms of use seemingly offer little leeway, asserting that user-generated content is subject to perpetual and irrevocable licensing to Stack Overflow for various purposes, including AI training.
However, at present, Stack Overflow users find themselves unable to opt out of AI training data usage
A purported email from Stack Overflow’s moderation team, shared by Ben, justified his suspension by asserting that once content is posted, it becomes part of the collective contribution, thereby limiting individual control over it.
Stack Overflow’s Subscriber Content section outlines this perpetual and irrevocable licensing, affirming the platform’s authority to utilize, distribute, and exploit such content, even if it’s removed by the user.
Despite efforts by users like Ben to invoke the European Union’s GDPR’s right to be forgotten, Stack Overflow may contest this based on exemptions for data necessary for freedom of expression and information.
As the standoff continues, legal implications loom large, with potential ramifications for privacy and ownership rights. However, at present, Stack Overflow users find themselves unable to opt out of AI training data usage, awaiting further clarity from the platform.
