How to Make Apps and Websites Remove Your Nonconsensual Nudes | WIRED
CommentLoader-
Save StorySave this story
CommentLoader-
Save StorySave this story
Starting on Tuesday, May 19, tech platforms have to provide a way for people to report nonconsensual intimate images and videos, or NCII, uploaded to their platforms. The new requirement is thanks to the Take It Down Act, a law backed by First Lady Melania Trump that passed last year with bipartisan support.
The Take It Down Act applies broadly to a range of apps and online services, including social media and gaming platforms, according to business guidance published by the Federal Trade Commission, which is responsible for enforcing platforms’ compliance with the law.
WIRED reached out to 14 companies that disclosed federal lobbying spending on the act to ask whether they believed they are subject to the law, and if so, how people could submit a takedown request.
Several company spokespeople were quick to say their companies supported the legislation, but they took additional time to explain how people could actually file a takedown request. One assured WIRED that their platform offered a secure reporting form, but did not provide a link to the form in question until after WIRED followed up multiple times. Two companies provided links to support pages that appear to have been updated after WIRED reached out. At least two host their forms on third-party websites, potentially making it difficult for people to search for the form themselves. Others said they did not plan on launching their reporting forms until the day the law went into effect. Companies were given a year between the passage of the act and the date it became enforceable to set up their takedown systems and develop request portals.
A spokesperson for T-Mobile said that the company supported the act, but that it doesn’t operate the types of online platforms the act applies to and that its core business is providing wireless and broadband services. (Broadband services are explicitly excluded in the act.)
Some companies did not respond to WIRED’s repeated outreach at all, including “First Buddy” Elon Musk’s X Corp., which garnered international scrutiny earlier this year after its AI chatbot Grok generated and posted thousands of nonconsensual images of women in various states of undress at users’ prompting. Proton AG and Verizon also did not respond.
The FTC did not respond to a request for comment. However, FTC guidelines say that platforms have to make it easy for people to submit removal requests. (Disclosure: The author of this story previously worked for the FTC.)
“The reporting piece of this is one of the most important pieces” of the Take It Down Act, says Jennifer King, a fellow at the Stanford University Institute for Human-Centered Artificial Intelligence. “And it's the thing that the companies I think often overlook.”
King points out that many people who might want to report nonconsensually shared nudes or deepfakes of themselves might be teenagers, who aren’t necessarily aware of their rights under the law or can easily understand the legalese written by compliance professionals included in these takedown request forms.
King says, “A lot of trouble with these types of reporting forms is that they don’t put any resources into testing them–they don’t test them with younger users, they probably don’t test them at all.”
How a Takedown Request Works
Regardless of the platform, the Take It Down Act has a few basic requirements for what information needs to be included in a takedown request, says James Grimmelmann, a law professor at the Cornell Law School and Cornell Tech.
At minimum, a takedown request has to include a way for the platform to locate the content, such as a link, and a statement explaining that it was not uploaded consensually. The request also has to include a signature either from the person depicted in the content or someone authorized to act on their behalf, and a way to contact that person.
Once someone submits a takedown request, a platform has up to 48 hours to determine whether it is valid. If it decides that it is, then it has to remove both the content reported and any identical copies.
Several larger platforms say they use an industry tool called StopNCII, which uses matching algorithms to identify abusive images and videos and is maintained by a British nonprofit. People can open cases directly on the tool’s website to add to what the tool flags. Reddit, TikTok, Snap, Microsoft Bing, and Meta’s social media platforms Facebook, Instagram, and Threads are all listed as participants on the tool’s website.
Though many major platforms have dedicated forms to help guide the submission process, Alejandro Cuevas, a postdoctoral fellow at Princeton’s Center for Information Technology Policy who studied the impact of initial passage of the law on deepfake communities, has observed that some sites only offer an email address for people to submit takedown requests.
Cuevas says in those cases, keeping good documentatio