In general, when a survivor comes to the NCOSE Law Center and shares that content depicting their abuse is circulating online, the content removal and takedown process requires hundreds of hours writing, sending, and following up with various web hosting providers and websites. And even still, this content does not come down easily. Even when websites and web hosting providers are put on notice that the content contains Child Sexual Abuse Material (CSAM) or depictions of sex trafficking, they are often reluctant to remove the content, refuse to accept responsibility, or ignore the requests altogether.   

Even Google, with the newly rolled out nonconsensual explicit or intimate personal image reporting portal, requires the reporting party to provide the full name of the individual in the content, the URL(s) of the webpages that are displaying the content, the Google Search result URL(s), as well as screenshot(s) of the content being reported. Although we understand it is critically important that Google can identify the content requested for removal, the process of taking screenshots is extremely triggering to survivors, as well as those assisting victims with the removal process.  It also requires the survivor to duplicate and increase circulation of their own abuse. The harm outweighs the potential usefulness of this requirement – and we urge Google to explore other ways to obtain the same result.

As part of the report, the survivor can indicate that the content in question was uploaded without consent and/or depicts sex trafficking or child sexual abuse material. Once the request is submitted, Google will review the content to determine if it violates their policies. In our experience, the results can take several days to yield a response—in the meantime, the survivor’s images remain easily accessible online, thereby leaving the survivor to suffer on-going trauma and vulnerability to sextortion, harassment, and other life-altering consequences like losing a job, getting kicked out of school, or losing friendships as people may not know the context of the images. And even then, it may be inexplicably determined by Google moderators that the content does not violate their terms and conditions (despite the survivor’s testimony).

How can Google better determine consent than the person who was harmed? Even law enforcement and other experts often can’t tell if an image or video depicts consensual acts as no one can see who else is behind the camera threatening them to look like “they enjoy it” or know what threat was used to coerce the creation of the video.

Notably, copyright infringements appear to be handled more swiftly and thoroughly by Google than image-based sexual abuse. 

As noted above, the process to report these images is extremely cumbersome, and it is very difficult for moderators to be able to know whether consent has been given. But we do want to commend Google on advances made in deduping violative nonconsensual images through hashing technology and for delisting violative urls. We understand that Google does try to detect and remove copied versions of this content in Search, but that modified images may evade detection (at least until more refined technology to detect alterations is created). Google also added an “explicit content filtering” checkbox feature their reporting form last year – which does mitigate the need for users to keep resubmitting removal requests.  

Image-based sexual abuse (IBSA) is a broad term that includes a multitude of harmful experiences, such as nonconsensual sharing of intimate sexual images (sometimes called “revenge porn”), pornography created using someone’s face without their knowledge (or deepfake pornography), nonconsensual recording or capture of intimate nude images (this can include so-called “upskirting” or surreptitious recordings in places such as restrooms and locker rooms via “spycams” etc.), and more.  

Often, such materials are used for sextortion for additional sexual abuses, including sex trafficking, or to advertise prostitution and sex trafficking victims. 

A 2017 nationwide study found that 1 in 8 American social media users have been targets of the distribution of sexually graphic images of individuals without their consent. Women were significantly more likely (about 1.7 times as likely) to have been targets of nonconsensually distributed pornography compared to men. 

Experiencing image-based sexual abuse can cause long-lasting trauma. The Cyber Civil Rights Initiative conducted a survey of victims of nonconsensual distribution of explicit photos in 2012. Of the victims surveyed, 93% reported that they had suffered significant emotional distress due to their pictures being posted online. This can include high levels of anxiety, PTSD, depression, feelings of shame and humiliation, as well as loss of trust and sexual agency. The risk of suicide is a very real issue for victims and there are many tragic stories of young people taking their own lives as a result of this type of online abuse. 

Improvements Google must pursue: 

1. Institute a survivor-centered reporting and removal process by immediately removing reported content and blocking the ability for it to resurface while they investigate the report. 
2. As a matter of policy, Google should honor removal requests of sexually explicit content unless the uploader can affirmatively prove consent. This would also lift the likely traumatic burden on Google moderators to try to determine whether the rape scene they’re looking at is real or simulated.
3. When age/consent cannot be affirmatively proven, not only should the Google Search result be removed and prevented, but Google should also require the platforms hosting the materials to remove them or risk negative page ranking consequences.
4. Make it easier for users to find Google’s reporting mechanism for nonconsensual sexually explicit materials.
5. Create different options to identify violative content in order to decrease the burden on survivors to screenshot their own abuse when making a report.
6. Strengthen the hashing system for content deemed to be criminal or nonconsensual to prevent it from resurfacing again and continue to work with other companies to strengthen the technology industry as a whole in responding to and preventing such content from surfacing (Google has created two tools to hash for nonconsensual images: Content Safety API and CSAI Match that they share with others free of charge!). 
7. Adjust its algorithms so that search terms like “white supremacy porn,” “forced porn,” “drugged porn,” “incest porn,” “teen porn,” and “choking porn,” etc., lead to resources and educational information instead of pornography websites depicting videos of sexual abuse.
8. As a policy matter, only allow pornographic websites which employ meaningful age and consent verification to surface results and utilize marketing tools via Google Search.