Twitter kept more than 40 images of child sexual abuse despite being banned

2023-06-06 13:20:07

Researchers at Stanford University in the United States have determined that Twitter continued to distribute images of child sexual abuseeven though they have been marked as prohibited for their content.

Specifically, it was the researchers from the Internet Observatory of this institution who came up with this issue during an analysis focused on child safety problems on various platforms.

From The Wall Street Journal, which includes some of the conclusions of this study, they indicate that these analysts informed Twitter about what they had discovered, a problem that was initially resolved in May.

The researchers found that from March 12 to May 20, Twitter hosted more than 40 images that had previously been labeled as child abuse material and that they found after investigating approximately 100,000 tweets.

From Twitter they claim to have “zero tolerance” towards child sexual exploitation and emphasize that they eliminate any material that shows or promotes these practices, such as multimedia content, text, illustrations or images.

The pedophile content was more than two months

The direct consequence of the distribution of this material is the immediate suspension and permanent of these accounts and that offenders “are prohibited from creating an account again”, this pedophile content would have managed to reach other users and remain on the social network for more than two months.

In fact, the researchers used a tool called PhotoDNA to detect this content, the same one that the platform admitted to using to detect and delete material related to the pedophilia.

In January, the Elon Musk-owned platform claimed to have suspended around 404,000 accounts related to children’s material, which represented a 112 percent growth in account blocking since November.

Due to the increase in access levels to its application programming interface (API), many researchers agree that it will become more and more difficult to detect errors like thiswhich will hinder future research.

In fact, the Stanford Internet Observatory has admitted to stopping using Twitter’s enterprise-grade API – which is priced at about $4,600 – due to new costs.