Purpose of the tool database

The tools database aims to compile, as far as possible, the digital tools used in the fight against child sexual abuse, in order to aid in the development of new tools, and increase usage of existing tools.

The database contains more than 90 different AI tools used to combat child sexual abuse. Our tools are carefully selected and reviewed before being added to the database.

Furthermore, all resources in the digital tools database have been labeled as to what their purpose is, what technology they employ, and in which crime phase they can be used. The sub-categories of these labels can all be found in the database. They can also be sorted based on a specific category or label.

The digital tools database can be a valuable resource for developers of AI who are interested in creating tools to combat child sexual abuse. By reviewing the different tools already available in the database, developers can gain insights into the types of AI techniques and technologies that are being used successfully in this field. They can also use the database to identify potential gaps or areas where existing tools could be improved or augmented with additional features. In addition, the detailed labelling and categorization of the resources in the database can help developers to quickly find tools that meet their specific needs and criteria, which can save time and effort in the development process. Overall, the digital tools database provides a valuable starting point and reference for developers who are interested in using AI to combat child sexual abuse, as well as a platform for collaboration and sharing of ideas and resources within this community.

Criteria:

In order for the database to have a clear delineation of which tools should be included, the criteria below were created.

• The tool should be able to be used to prevent or combat child sexual abuse, or alternatively provide support after abuse has occurred.
  • Tools that can facilitate actors, but are not linked to child abuse, should not be included. For example, the translation tool DeepL should not be included, as it is not linked to child abuse, although investigators may use it during investigations.
• The tool should be of a digital nature, such as an application, platform or database.
• The tool should contain a reasonably advanced technical component, which should add something to the tool.
  • Examples of reasonably advanced or advanced technical components: AI classifying images, search engines, simple prioritisation algorithms, databases with hash lists, a chatbot whose responses are generated by machine learning models.
  • Examples of non-advanced technical components: a website with articles, a chatbot with pre-written answers and conversation flows, an app with information texts sorted into categories.