X owner Elon Musk has announced that any censorship required by governments on the platform is now clearly visible to users.
In a post on August 15, Musk wrote: “Any censorship required by governments is now clearly visible.”
The statement came in response to an update from X’s open-source algorithm team, which detailed the latest changes to the “For You” ranking system, including a new filter related to Brazil’s 2026 election.
Under the approach, users will be informed when a government authority seeks the removal of a specific post or the restriction of an account.
X also plans to identify the government agency involved and, where possible, disclose the legal grounds or explanation provided for the request.
Any censorship required by governments is now clearly visible https://t.co/eQMdoYlhkA
— Elon Musk (@elonmusk) August 15, 2026
X’s open-source algorithm repository on GitHub was updated on August 14 to detail the latest changes to the “For You” recommendation feed. The code has been published under an Apache v2 licence and covers model configuration, filters and core ranking systems, expanding the company’s earlier open-source release by roughly 10 to 15 times.
The release explained how ranking weights function and introduced a new filter required specifically for Brazil’s 2026 election. The repository states that the Brazil2026ElectionFilter removes posts by accounts reported to the country’s Electoral Court from the For You feed unless the viewer follows the account, citing Brazilian electoral law. Brazilians vote in the first round of general elections on October 4.
X vice-president of product Keith Coleman told TechCrunch that outside developers would be able to submit changes through pull requests, and that the aim was that the code “can be audited; it can be critiqued”. The company said it had withheld systems that use Grok to predict rule-breaking posts, to stop bad actors evading enforcement.
X is also adding an “Under the Hood” settings page where accounts that have posted 10 or more times in the past month can download data showing whether labels were applied to them, starting with a pilot group of accounts at least one year old.
The move forms part of Musk’s broader push for greater transparency around content moderation and recommendation systems on the platform.
X has previously published transparency reports detailing government requests for information and content removal, and participates in public databases that archive such notices.
Crypto Briefing has reported that under Musk’s ownership X has complied with a high proportion of government takedown requests, with figures ranging from 83 per cent to 98.8 per cent depending on the reporting period, a rate higher than that recorded under the platform’s previous management.
Musk has repeatedly stated that X adheres to the laws of the countries in which it operates, while arguing that users should have clear visibility into when restrictions are imposed by state authorities rather than by the platform itself.
Before Musk’s acquisition of Twitter in 2022, the platform faced widespread accusations of systematic bias against conservative voices.
Internal practices included “visibility filtering”, often described by critics as shadowbanning, which limited the reach of accounts without notifying users.
Prominent conservative figures such as radio host Dan Bongino, Turning Point USA founder Charlie Kirk and Stanford professor Jay Bhattacharya (a critic of Covid-19 lockdowns) were placed on internal blacklists that prevented their posts from trending or appearing in search results. Screenshots published by journalist Bari Weiss showed Bhattacharya tagged “Trends Blacklist”, Bongino “Search Blacklist” and Kirk “Do Not Amplify”. The platform also permanently suspended then-President Donald Trump’s account in January 2021.
These practices came under intense scrutiny after Musk granted selected journalists access to internal documents, resulting in the “Twitter Files” series published from December 2022. The releases detailed blacklists, algorithmic suppression and coordination with government agencies and political campaigns on content moderation decisions.
The files brought public attention to previously opaque processes that Twitter executives had repeatedly denied involved viewpoint-based censorship.
In Europe, the European Commission has repeatedly pressed X for greater transparency over its algorithms and content moderation systems, ordering the platform in January 2025 to hand over internal documents on recommendation mechanisms and fining it €120 million on December 5, 2025 for breaches of transparency rules on advertising, data access for researchers and verification badges. It was the first non-compliance decision issued under the Digital Services Act (DSA), and X, Musk and xAI Holdings lodged appeals against it at the EU General Court in Luxembourg on February 16.
The Commission extended its December 2023 proceedings against X on January 26 to examine how the platform manages risks tied to its recommender systems, including its shift to Grok-based recommendations.
At the same time, the DSA imposes obligations on very large online platforms to assess and mitigate systemic risks, including the spread of illegal content and disinformation.
Critics of the regulation, including Musk, have argued that these requirements effectively pressure platforms into proactive censorship, while the EU maintains that the rules are content-neutral and focused on accountability.
