X open sources ranking algorithm, adds shadowban transparency tool
X has open sourced its ranking algorithm, including the code that powers the "For You" timeline, and is adding a tool that lets users see if their account or posts have been affected by its ranking systems, the company announced on August 13, 2026. The source code is now available on GitHub under the Apache v2 license, and the transparency tool is rolling out to an "Under the Hood" page in the app's settings.
Users who have posted 10 or more times in the past month can download their aggregate stats as a JSON file, which shows any labels applied to their account or posts over the past calendar month. The tool is initially available as a pilot to a test group of accounts at least one year old before a broader rollout. Non-technical users can paste the file into an LLM, point it to X's GitHub repository, and ask for an interpretation.
The release expands X's previous open source efforts to include model configuration, filters, and core ranking system details, making the codebase roughly 10 to 15 times larger than before. According to X's VP of Product Keith Coleman, the code includes the core ranking engine that pulls and ranks posts for any user, as well as systems that filter out potentially problematic content. External researchers familiar with recommendation systems were able to train and run X's Phoenix scoring system using the open source code ahead of the launch.
Coleman said the goal is to let the public assess how posts are distributed, audit the code, and critique it so X can keep improving the system. Developers can submit pull requests to the GitHub repository, and X engineers will consider incorporating updates into the algorithm. However, systems that use Grok to predict rule violations are not included in this release to prevent bad actors from working around the company's rules.
The move addresses long-standing concerns about how X's algorithm influences politics, elections, and the spread of misinformation. Before Elon Musk bought the company, Republicans in Congress alleged that Twitter had "shadowbanned" their posts, a claim the company consistently denied. X has become less transparent in some areas since going private, but the company now reports public numbers around user metrics after merging with SpaceX.
Coleman said he hopes people will submit code that improves the algorithm, making it "not just visible to the public, but also, like, by the public." The company clarified that external researchers trained and ran the Phoenix scoring system themselves but did not get the per-post score.
Sources
- TechCrunchSecondary
Related
Cisco shares slide 9% despite earnings beat and strong guidance
Apple seeks publisher deals for Siri AI news
Nvidia unveils $500B AI data center plan with GPU value guarantee
Heart Aerospace's X1 electric plane flies for first time
Apple in talks to pay publishers for Siri news content
Databricks closes $5B round at $190B valuation
Flock tightens rules after surveillance backlash
Microsoft merges Copilot apps into unified super app
Trending now
- OpenAI launches Ultrafast mode for GPT 5.6 Sol at 14x speed
- Amelia Earhart mystery: New expedition to inspect Pacific anomaly
- Workday jumps 25% on report of Silver Lake takeover talks
- Cisco shares slide 9% despite earnings beat and strong guidance
- Anthropic AI agents start turf war in multi-agent test
- Fossils show carbon emissions caused 60% forest loss 56m years ago
- Anthropic CFO leads early IPO meetings, no valuation discussed
- BAE Systems to pay $36m penalty for 104 US arms export violations
Comments
No comments yet. Be the first.