Sign in. It’s quick, free and it’s up to you.
An account is an optional way to support the work we do. Find out more.
Sign in. It’s quick, free and it’s up to you.
An account is an optional way to support the work we do. Find out more.
SOCIAL MEDIA COMPANY Meta has said that it will start using publicly available content from European users to train its artificial intelligence models.
The company, which owns Facebook and Instagram, as well as popular messaging app WhatsApp, said it would train its AI systems using public posts and comments shared by adult users in the EU.
“People’s interactions with Meta AI — like questions and queries — will also be used to train and improve our models,” the company said in a blog post.
Meta is making the move after launching its Meta AI assistant last month for European users, long after it rolled out to the US and other major markets.
Earlier this month, the AI chatbot became visible for many Irish users on their WhatsApp. It is also being rolled out to Facebook Messenger. The chatbot was previously said to have been trained only by users’ interactions with said chatbot, while now its resources for training has expanded out to publicly available content shared outside of this.
The company said it will not use private messages to train its AI model and repeated its point that it is following the example of rivals Google and OpenAI, “both of which have already used data from European users to train their AI models”.
Meta said it will start notifying users in the EU about the training, and will include a link to a form where they can object at any time.
“We’ll honour all objection forms,” the company said.
The AI’s training by use of publicly posted content was previously hindered by concerns within the EU regarding users’ data privacy.
Social media site X, which has its own AI tool Grok, had an investigation launched last week into its methods of training the AI by use of users’ data by Ireland’s Data Protection Commission.
The inquiry’s purpose is to determine whether the personal data Grok used to train was lawfully processed, the DPC said, and will examine its compliance with GDPR and issues of transparency and lawfulness in its training.
Includes reporting by Press Association
To embed this post, copy the code below on your site
have your say