Indonesia lets Elon Musk’s Grok back online under tight supervision

Indonesia allowed Elon Musk’s artificial-intelligence chatbot Grok to resume operations in the country on Feb. 2, the communications ministry said, describing the authorization as conditional and subject to close oversight.

The ministry said X Corp. made a written commitment to service improvements and to comply with applicable laws, according to a statement shared with the ministry.

In the same statement, the ministry said the company told officials it had taken steps to address misuse of Grok services, including restricting access to certain features.

The decision came weeks after Indonesia banned Grok for explicit sexual content, following concerns that it had been misused to generate sexually explicit and nonconsensual images.

Indonesia said it will not treat the return of Grok as an open-ended reopening. Alexander Sabar, the ministry’s director general of digital space supervision, said Indonesian authorities would verify and test the company’s claimed measures to ensure they prevent violations, including the distribution of illegal content and breaches of child protection principles.

“If inconsistencies or further violations are found in its implementation, the Ministry of Communication and Digital Affairs will not hesitate to take corrective action, including suspending access to services again,” Sabar said.

Indonesia also pointed to regional developments in its explanation. It said Indonesia and Malaysia were the first two countries that blocked access to Grok in January over concerns about misuse for sexually explicit and nonconsensual images.

Malaysia later lifted its temporary restriction after X took “security and preventive measures,” and Malaysia’s regulators said they met with X representatives and would continue monitoring.

The Indonesian ministry’s statement did not outline specific testing timelines or the technical scope of the feature restrictions, but it said the steps X cited would be checked against compliance and child-protection standards, with access potentially halted again if problems persist.