Twitter is asking users how it should tackle AI-generated "deepfake" videos and edited images designed to mislead the public.
The company has issued a survey, the results of which will be used to shape policy concerning "synthetic and manipulated media," amid fears the content will be used to spread misinformation during the 2020 presidential election.
"When you come to Twitter to see what's happening in the world, we want you to have context about the content you're seeing and engaging with," Twitter's VP of Trust and Safety, Del Harvey, said in a blog post.
The company defines synthetic and manipulated media "as any photo, audio, or video that has been significantly altered or fabricated in a way that intends to mislead people or changes its original meaning."
Currently, the company is considering four different actions to address tweets containing manipulated media. They include placing a notice next to them about how the content's been altered, warning people before they share the content, and adding a link to a news article explaining why the media has been fabricated. The most drastic action involves deleting the tweets when it's evident the misinformation could harm someone's safety.
In the survey, users can weigh in on how far the social media company should crack down on the content. Should it remove the posts? Or basically do nothing? Another question asks whether it would be acceptable for Twitter to effectively hide tweets that contain the content, making them harder to find on the platform. Or if the company should alert people when they share them, thereby forcing users to think twice about circulating the content.
The survey also asks under what circumstances should Twitter remove a tweet that contains manipulated media. Most people would agree a deepfake video that risks harming a person's physical safety should be deleted. But what if it only risks damaging a person's or group's reputation, privacy, or just their financial wealth? Also, what about their mental health?
Twitter is accepting the feedback from the survey until Nov. 27. When the company does enact the policy, Twitter will announce the plan at least 30 days before it goes into effect.
Users can also tweet their feedback with the hashtag #TwitterPolicyFeedback.