Facebook’s Meta announced this week the release of a new tool that uses artificial intelligence (AI) to identify web content that may contain it. information incorrect various arguments. sphereIt uses a database of more than 134 million articles to feed the platforms it supports, through an all-read open source system. check for fake news.

New AI-powered Meta tool (formerly Facebook) lets you report fake news and disinformation

Based on how much was delivered by the company, one of the first services to use IA, WikipediaThe encyclopedia is about the cooperation of the best people in the world for the arguments of various generations. According to Meta, Sphere will be used by the editors team. As you can see from the Developer of the site to make it easier for users to control the proposed changes, the AI ​​tool can be a powerful ally in the war on disinformation by compiling the information of web pages. All for the purpose of facilitating the verification of texts containing erroneous quotations or extra information.

Initially, these sources should be geared towards B2B-only to combat fake news, given data from a number of reliable sources. rather than using a unique site or collection as a primary source checking the recommendations, making it more complete than other options available on the website.

So far, la Wikimedia BasisThe company that owns Wikipedia, It is the only Meta client that uses Sphere to identify errors in article suggestions. However, if the service improves, more developers and companies will need to adopt AI services.

Automated tools can help identify inconsistencies or special wording of citations, but they only help real editors determine if a source is really a much more complex artificial analysis that requires the depth and depth of a system of understanding. The team developed this based on research and progress in Meta AI The first model in the degree of automatic expansion of hundreds of copies of contemporary quotations to verify that respondents’ claims are fully supported.

Shani Evenstein SigalovProfessor and researcher at Tel Aviv University and vice chair of the Wikimedia Foundation’s founding council said:

This is a powerful example of machine learning tools that can help scale volunteers’ work by providing efficient citations and accurate resources. Improving these processes will allow you to attract new editors to Wikipedia and provide better and more reliable information to billions of people. I don’t see time to continue to evolve in this area, especially since the machine learning tools are at a level to provide more personalized citations and multilingual options to serve our Wikimedia community in over 300 languages.