Meanwhile, Wikipedia opens up to more AI innovations
Meta, the company behind Facebook, is to provide more personalized ads using the conversations users have with their AI assistants (both audio and text).
The policy will be implemented worldwide from December 16th and could affect up to 1 billion people worldwide, but users in the UK, European Union and South Korea will be excluded based on data privacy and protection grounds. The goal is probably to start driving revenue from AI. This is an expensive service that takes into account the IT overhead required for support.
“Chat with Meta AI about hiking and you will see that you are interested in hiking just as you would if you posted a reel about hiking or if you like hiking related pages. As a result, you may start seeing recommendations for hiking groups, friend posts about trails, or ads for hiking boots.
Affected users will be notified on October 7th.
Data extracted from Meta AI conversations will be added to existing targeted ad databases that contain information about people's obvious likes and dislikes. However, the company argues that it does not offer targeted advertising based on sensitive topics such as sexual orientation, political views, health, racial and ethnic origins.
This shift coincides with founder and CEO Mark Zuckerberg's “major individual AI focused on personalization, voice conversation and entertainment.”
However, it is open to question whether Meta AI can currently compete with ChatGpt or Grok. When asked to provide Meta AI with more information about the initiative, they were unable to provide a source of information from the company, but added, “Unless you do not use Meta's AI tools, users have no option to opt out of this data collection.”
Meta is not the first company to begin deploying AI with AD Personalize services. While Google and Amazon have already gotten off to a start via cloud-based services, Meta is the first to sign up for AI conversations with users with a service that drives advertising revenue.
Wikipedia's AI Movement
At the same time, Wikipedia officially launched a new project that could open up open source encyclopedias to AI models.
Called the Wikidata Embedding Project, it is led by Wikimedia Deutschland in collaboration with Jina.ai and DataStax. “This project aims to use Wikidata's multilingual, comprehensive knowledge graphs to support the open source community in the development of innovative AI applications, making the broader data more accessible and contextually relevant for users around the world.”
The project was launched in 2024, but officially launched on October 1, 2025.
Although organizations have provided machine-readable data from their properties for many years, information can only be misused in the form of keyword matching, and you may miss the actual meaning behind search queries.
“SPARQL [Wikidata’s Query Service] It provides accurate and accurate data search, but its sharp learning curve and complexity make it difficult for many users to take advantage of. Vector base [semantic search] The approach bridges this gap by combining keyword search accessibility with context-enabled results,” the organization explained.
The database is now accessible through the open source website Toolforge.
