A good legal question is whether the selection/rejection of content by an algorithm tuned to provide financial benefit to the platform would be considered "editorial control".
If there were a dead simple filter where the user could pick friend groups, tags and sorting criteria to tune their feed this would not be an issue. Reddit, for instance seems relatively simple in that respect- the presentation is a function of the subreddit and the votes.
But once the algorithm is driven by sponsorships, monetization opportunities, and opaque surveillance data then the control of the presentation shifts from user to platform. One could argue that this should creates some liability on the part of the platform.
If the editor of a publication was monetarily compensated for publishing lucrative slander, in a manner designed to maximize its credibility with certain audiences, and it resulted in harm to people then they arguable could be held responsible. If the editor claimed that an algorithm decided to publish it and the publication developed the algorithm I don't think it would make them less responsible.
If there were a dead simple filter where the user could pick friend groups, tags and sorting criteria to tune their feed this would not be an issue. Reddit, for instance seems relatively simple in that respect- the presentation is a function of the subreddit and the votes.
But once the algorithm is driven by sponsorships, monetization opportunities, and opaque surveillance data then the control of the presentation shifts from user to platform. One could argue that this should creates some liability on the part of the platform.
If the editor of a publication was monetarily compensated for publishing lucrative slander, in a manner designed to maximize its credibility with certain audiences, and it resulted in harm to people then they arguable could be held responsible. If the editor claimed that an algorithm decided to publish it and the publication developed the algorithm I don't think it would make them less responsible.