As concerns regarding AI-generated false images escalate, Twitter is seeking to leverage the power of its users to combat this issue. The social media giant recently announced an expansion of its Community Notes feature, which enables crowdsourced fact-checking. This innovative feature now encompasses image verification as well. The move comes after Twitter circulated an AI-generated image falsely depicting a bomb at the Pentagon, causing widespread alarm.
The responsibility of identifying misleading media now falls into the hands of Twitter users, a drastic shift from the company’s original moderation strategy. Since Elon Musk’s takeover as owner in October, Twitter’s trust and safety team, previously tasked with identifying fabricated or misleading content, has significantly diminished. The company has not expressed intentions to recruit additional content moderators to check the proliferation of deceptive images.
Community Notes’ broader scope now permits users with an impact score of 10 or higher to append notes to a specific image within a tweet. Impact scores essentially reflect the effectiveness of a user’s contributions to the platform. Users can employ these notes to provide more context to an image or flag content that might be manipulated or AI-generated. Twitter affirms that these notes will automatically accompany “recent and future matching images”, indicating that a potentially questionable image will retain the note even when republished by other users on the platform. However, Twitter’s confidence in its ability to align tagged images with all other similar versions appearing on the site seems somewhat lacking at the moment.
Twitter, citing an aim for precision in image matching, acknowledges that the system might not identify every visually similar image. The company assured that efforts are underway to refine the system to maximize coverage without yielding false matches.
Currently, the pilot feature only supports single-image tweets. Nevertheless, Twitter expressed ambitions to broaden this service to include videos and multi-media tweets. When Gizmodo sought more details about the program, the company’s response was notably evasive, delivering only a poop emoji.
The adoption of a frugal approach by Musk, particularly concerning Twitter’s API access, has effectively alienated a host of beneficial third-party apps. These apps were instrumental in combating some of the most harmful content on the platform. Block Party, a highly regarded anti-harassment tool that served to block and mute countless trolls, announced its indefinite hiatus from the platform on the same day Twitter expanded Community Notes.
Block Party expressed its regret over the situation, stating, “We’re heartbroken that we won’t be able to help protect you from harassers and spammers on the platform, at least for now; we fought very hard to stay, and we’re so sorry that we couldn’t make it happen.”
