Friday, June 14, 2024, 7:37 p.m.
| Updated 7:59 p.m.
Where I said I say, I say Diego. This has been done by Meta, Mark Zuckerberg’s conglomerate, days after confirming and reaffirming that he would use images from Facebook and Instagram to train his artificial intelligence (AI) algorithms. Now, a few weeks later, it is backing down and will not use the photographs that its European users upload to its platform in order to comply with European Union regulations.
In a statement, made public this Friday, the social media giant has defended its technology and its commitment to training its intelligent tools. “We are confident that we comply with European laws and regulations.” Four brief paragraphs that have been “welcomed” by the Irish Data Protection Agency (DPC): “This decision has been taken after intense negotiation between the DPC and Meta.”
The decision by the parent company of Facebook and Instagram to pause the training of its AI algorithms in Europe comes after the European Center for Digital Rights, a privacy advocacy group based in Vienna, Austria, filed 11 complaints due to a change in the privacy policy.
Last May, this newspaper announced that Meta had planned to introduce modifications to its terms to be able to use users’ publications, audios and photographs so that its AI can analyze and be able to generate information. “This would allow the illegal use” of personal data to train an undefined type of current and future artificial intelligence technology,” spokespeople for the Austrian group alleged earlier this week.
A controversial change
With a brief notification, Meta notified its millions of users of this ‘small’ change a few weeks ago. Many others directly realized it when they saw posts from their friends on Facebook and Instagram.
«I own the rights to all images and posts published on my Instagram profile and therefore do not give my consent for Meta or other companies to use them to train generative AI platforms. This includes all future and past posts/stories/threads on my profile. @Instagram get rid of the AI program”, began to be read in Meta stories and posts.
An advertisement that generated a lot of rejection among the users most jealous of their privacy, who refuse to let a private company train their models based on their texts and images. «We are preparing to expand our AI at Meta experiences [sus herramientas de IA generativa] in your region. (…) In order to bring you these experiences, we will rely on the legal basis called legitimate interests to use your information in order to develop and improve AI in Meta,” argued the company owned by Mark Zuckerberg.
However, this change, at least for Facebook and Instagram users in Europe, is on hold. “We are disappointed by the request from the Irish Data Protection Commission (DPC), our lead regulator, on behalf of European DPAs, to delay the training of our large language models (LLM) using public content shared by adults on Facebook and Instagram, especially as we have incorporated regulatory comments and European DPAs have been informed since March. This is a setback for European innovation, competition in the development of AI and further delays the benefit of AI for people in Europe,” warns the American technology company.
€2/month for 4 months
Are you already a subscriber? Log in
