Users on social media platform X can reportedly choose whether or not their posts will be used to train Elon Musk's Grok AI, as revealed by the company.
(Photo : BoliviaInteligente from Unsplash)
In March, X, former Twitter was already teasing its plans to allow NSFW content on the platform. It turns out that it's now officially allowed to be posted.
By default, X users' settings permit the chatbot to access their personal information. However, as stated by the firm in a post on X, all X users can decide whether or not their publicly visible postings can be used to train Grok, the AI search helper. This setting will soon be made available on mobile devices in addition to the web platform.
The business also mentioned that customers have the option to make their accounts private to shield their data from the chatbot.
As a competitor to OpenAI's ChatGPT, the artificial intelligence chatbot Which is presently accessible to X premium subscribers introduced in November of last year. It is a product of the startup xAI, which Musk also owns, and was trained using data from X.
Meta Faces Legal Scrutiny
The opt-in and out option on X comes after Meta, who announced back in June that it will be training its AI model on Facebook and Instagram posts garnered legal scrutiny from several governments.
More precisely, the national data protection body of Brazil recently ruled that Meta could not use data mining in the nation to train its AI models.
The decision, which was published in the nation's official gazette, reportedly forbids Meta from feeding data from Brazil into its AI systems. The regulatory body stated that concerns about possible harm to people's fundamental rights led to this conclusion.
Brazil is a significant market for Meta, with Facebook alone boasting about 102 million regular users out of a population of about 203 million.
Despite Meta's claim that its revised privacy policy conforms with Brazilian privacy rules, the firm voiced dissatisfaction with the decision, pointing out how it would affect innovation and the advancement of AI in Brazil.
Challenges have been raised by Meta's updated privacy policy in Europe as well as Brazil, where intentions to include public posts in AI training were recently shelved.
On the other hand, sources claim that comparable procedures are already in place in the US, where federal privacy regulations are less strict. In previous announcements, Meta stated that it intended to improve AI capabilities by utilizing publicly available information about its products and services, including data from publicly posted images and captioned photos.
For users who would prefer not to have their data used in this manner, questions have been raised over the opt-out process's accessibility and transparency. The regulatory ruling also takes into account more general worries regarding the use of private information in AI research.
Problematic AI Training Data
Human Rights Watch recently reported on cases in which identifiable Brazilian children's private images were taken from several internet platforms without their express agreement and used to build AI tools. In some cases, this resulted in the production of explicit material created by the AI.
Critics contend that the ruling may discourage businesses from openly revealing their data usage policies, even though the decision has been commended for its ability to uphold privacy rights.
Related Article : X Private Likes Fiasco Made Public was Because of 'Security Incident,' Company Claims
