Google has announced a partnership with StopNCII.org aimed at tackling the spread of non-consensual intimate imagery (NCII) online. The company confirmed on Wednesday that, over the coming months, it will begin using hashes provided by StopNCII to detect and remove such content from its search results proactively. Hashes are unique identifiers generated through algorithms that allow services to identify abusive imagery without sharing or storing the original material. StopNCII uses PDQ hashes for images and MD5 hashes for videos.
Acknowledging past delays
Google’s announcement comes amid criticism that it has been slower than other tech firms in adopting such measures. “We have also heard from survivors and advocates that given the scale of the open web, there’s more to be done to reduce the burden on those who are affected by it,” the company stated in its blog post. Social media platforms, including Facebook, Instagram, TikTok, and Bumble, joined StopNCII as early as 2022, while Microsoft incorporated the system into Bing in September last year. Google’s move is therefore seen as a step towards aligning with industry standards, albeit later than some competitors.
Current tools and ongoing challenges
Previously, Google introduced tools that allow users to request the removal of intimate imagery, as well as personal contact information. However, like earlier efforts to address revenge porn, these initiatives placed the responsibility on victims to identify and flag the content themselves. Advocates have highlighted the difficulty of addressing AI-generated content or material that victims cannot easily locate, urging Google to explore solutions that do not rely solely on victims to submit hashes. Tackling these challenges remains a complex task, but experts say such measures are crucial in reducing the harm caused by non-consensual content.
Google’s use of StopNCII hashes marks a significant move to better protect individuals from online abuse better and brings the company closer in line with other major technology platforms.
