Tech giants are seeking to dominate the new world of artificial intelligence. In their quest, they’re causing controversy, raising concerns over user privacy and more. Adobe’s recently updated terms of service angered and worried some users when they were announced on the week of June 3.
The updated terms reportedly allow the company to have free access to users’ projects and Adobe Cloud. After backlash on social media, Adobe said that it will access user content only for purposes like taking action against prohibited materials and for innovation.
The innovation caveat led to more pushback from those users concerned about their work being used potentially for AI purposes. The concerns forced the company to respond again, saying that it “does not train Firefly Gen A Models on customer content” and “will never assume ownership of a customer’s work.
Meta, on the other hand, isn’t going the same route as Adobe. The company said that it will be using individuals’ personal posts, pictures and private messages on Facebook, Instagram and WhatsApp to help train its AI products. Meta said it doesn’t need users’ consent.
The controversial move from the Meta platform and other companies comes as a report on Thursday showed that companies will exhaust available human data for training purposes for AI language models sometime between 2026 and 2032. The reason is because the current pace of data mining won’t allow for new blogs, new articles and new social media content to keep the same pace into the next decade.
With Meta’s announcement, privacy advocates have sounded the alarm. A European advocacy group filed complaints on behalf of 11 countries against Meta for its new data-scraping policy. The group called on European Union authorities to halt the plan by the social media giant.
However, Meta argued that it doesn’t need permission from users. The tech giant claims that its new policy, which goes into effect on June 26, follows EU privacy laws.
Meanwhile, European users have the option to opt out of the policy while American users do not. In the United States, Meta has already gained access to public user data on Instagram, WhatsApp and Facebook.
As the “gold rush” for human-driven data continue by Big Tech, OpenAI and Google have reached contracts with Reddit and some news media outlets to train AI language models.
Meanwhile, several U.S. newspapers are fighting back against the data grab by Big Tech by suing OpenAI and Microsoft Corp. The newspapers are accusing the companies of stealing copyrighted articles without attribution to train AI systems.
