Twitter is cracking down on synthetic and manipulated media.
Based on input from global users, the social network updated its rules regarding “misleading” content—i.e. deepfakes.
The new decree, announced this week, states that no one shall “deceptively share synthetic or manipulated media that are likely to cause harm.”
Twitter also reserves the right to label posts containing such material, “to help people understand [its] authenticity and to provide additional context.”
“People recognize the threat that misleading altered media poses and want Twitter to do something about it,” Head of Site Integrity Yoel Roth and Group Product Manager Ashita Achuthan wrote in a Tuesday blog.
The post details criteria for the labeling or removal of tweets:
- Has it been substantially edited to add or erase visual or auditory information?
- Can the context in which it is shared result in confusion, misunderstanding, or suggestion of deliberate intent to deceive?
- Does it feature threats to physical safety, privacy, or the ability of someone to freely express themselves?
Twitter will use this criteria to consider media for labeling or removal under its rule (via Twitter)
Facebook has taken a similar response to manipulated media, promising to remove “misleading” content that has been edited or synthesized using artificial intelligence or machine learning in ways an average user can not easily spot.
The policy, however, does not cover all doctored videos: Content deemed parody or satire is excused, as are clips edited solely to omit or change the order of words.
Last year’s viral video of Speaker Nancy Pelosi—modified to make her appear drunk during a public appearance—does not appear to fit Facebook’s new guidelines.
Neither does a recent recording of former VP and presidential hopeful Joe Biden, which was heavily trimmed to make him sound racist.
Twitter, meanwhile, specifically considers whether media depicting a real person has been fabricated or simulated.
Users were nearly unanimous in their support for Twitter providing additional information or context on misleading altered media: Suggestions include placing warning labels next to significantly altered content and alerting folks before they share misleading media.
Compared to waving red flags, respondents were “somewhat less supportive” of removing or hiding misleading altered media, citing free expression and censorship.
Twitter will start labeling tweets with this type of media on March 5.
“This will be a challenge and we will make errors along the way,” the company said. “However, we’re committed to doing this right. Updating our rules in public and with democratic participation will continue to be core to our approach.”
More on Geek.com:
- Watch: Deepfake Hollywood Roundtable Gets Weird
- Deepfake Tool Makes It Easy to Put Words Into Someone’s Mouth
- Jack Dorsey: Twitter Will ‘Probably Never’ Get Edit Button
from Geek.com https://ift.tt/31u1Vm0
via IFTTT
0 comments:
Post a Comment