Very large data sets are needed to train artificial intelligence. The larger the dataset, the better the artificial intelligence performs. LAION It also creates datasets for artificial intelligence developers. Stanford Internet Observatory, LAION-5B It revealed hundreds of links to child abuse in the dataset.

LAION-5B, creator of stable diffusion Stability AI It was also used by. Researchers at Stanford, who began examining the dataset in September 2023, tried to find out whether there was child abuse in this dataset and, if so, how many there were. At least according to the research results 1679 contents Links were found to images depicting child abuse. This information was also shared with institutions such as PhotoDNA and the Canadian Child Protection Centre.

This data set was also used for stable diffusion

According to the information on the LAION website, the dataset does not store images, but creates an Internet index with text descriptions of the images and links to the images. At Google, Imogen is an older version of LAION-5B for training generative artificial intelligence. LAION-400M had used it. While the company said that 400M was not used in later versions. Imogen researchers are also in the dataset “There is a lot of inappropriate content, including child abuse, racist comments and harmful social stereotypes.”he claimed.

Stanford researchers say the existence of this content has no direct influence on the output of the dataset While LAION stated that they had a zero-tolerance policy against such harmful content and that they would temporarily withdraw the dataset from publication. On the other hand, retraining artificial intelligence trained on this data poses a bigger problem.

Previously, US prosecutors have presented to Congress the use of artificial intelligence in child abuse with productive artificial intelligence He called for the convening of a committee to prevent the production of such content.