Fazry Ismail / EPA
Grok
Businessman announces that the social network will be open source from Saturday. Indonesia and Malaysia were the first countries to block AI assistants.
The tycoon and owner of XElon Musk, announced that, starting next Saturday, his social network will be open source, including all code used to determine the organic and advertising posts that are recommended to users.
This measure comes after a series of controversies that involved the platform related to the generation of images explicit with artificial intelligence (IA) at the request of users, including nudity of minors.
After complaints from users and institutions, such as the European Union, which ordered the platform to keep all documentation until the end of the year, Musk took the decision to make the new algorithm public to provide more transparency to the social network.
In a message published on X, Elon Musk clarified that this measure “will be repeated every four weekswith full notes for developers to help them see what has changed.”
However, it’s not the first time that the tycoon is trying to make the open source of After that, the platform’s engineers argued that it was for “tracking statistics”, although they later eliminated that section of the algorithm.
Grok blocked
Indonesia e Malaysia were the first countries, this weekend, to block the Artificial Intelligence (AI) ‘assistant’ Grok, from the company xAI owned by North American Elon Musk, because they consider the tool to generate images of non-consensual sexual content.
The “chatbot” (contact robot) Grok, through the digital platform on the X network, has been targeted for creating manipulated images, including representations of women of bikini or in sexually explicit poses, as well as images involving children.
The regulatory bodies of those two Southeast Asian countries declared that the absence of control mechanisms to prevent the production and dissemination of false pornographic content. The Indonesian government temporarily blocked access to Grok on Saturday, followed by Malaysia on Sunday.
“The Government considers non-consensual sexual ‘deepfakes’ (artificial content) to be a serious violation of human rights, dignity and security of citizens in the digital space,” said Indonesia’s Minister of Communications and Digital Affairs, Meutya Hafid, in a statement.
These measures reflect growing global concern about generative AI tools that can produce realistic images, sounds and text, while existing safeguards fail to prevent their misuse.
The general director of supervision of the digital space, Alexander Sabar, in another statement, stated that it was proven that the Grok has no effective safeguards to prevent users from creating and distributing content pornographic based on real photos of Indonesian citizens, adding that such practices violate privacy and image rights, and can cause psychological, social and reputational damage.
In Kuala Lumpur, the Malaysian Communications and Multimedia Commission ordered a temporary restriction on Grok on Sunday, following what it called “repeated misuse” of the tool to generate obscene, sexually explicit and manipulated images without consent, including content involving women and minors.
The local regulatory body said that notifications issued this month to Musk’s companies (X Corp. and xAI), demanding more robust safeguards, received responses based mainly on user complaints.
“The restriction is imposed as a preventative and proportionate measure while legal and regulatory processes are ongoing,” declared the business group, adding that access will be blocked until effective safeguards are in place.
Launched in 2023, Grok is free to use on the X social network. Subscribers can ask questions on the platform and tag posts that they created directly or responses to posts from other users.
Over the summer, the company added an image generator feature, the Grok Imagine, which included a so-called “spicy mode” that can generate adult content.
The restrictions in Southeast Asia come as there is a growing scrutiny to Grok in other countries, including European Union, United Kingdom and India.
Last week, the generation and editing of subscriber images was limited following a global backlash over sexualized ‘deepfakes’ of people, but many experts argue that the measure did not completely solve the problem.
