X Open Sources Algorithm Amid Transparency Push and Grok Scrutiny

X has open sourced its algorithm once again. The move follows a promise made by Elon Musk last week. As a result, transparency has returned to the spotlight. The company shared its feed code on GitHub.
It also included a clear diagram explaining how recommendations work.
However, critics say the details remain limited. The algorithm reviews user activity and recent posts. It also scans content from accounts users do not follow.
Therefore, recommendations mix familiar and new material.
Certain posts are filtered out automatically. These include blocked accounts, muted terms, and spam-like content.
In addition, violent material is removed.

AI, Grok, and Ongoing Concerns

X confirmed the system relies fully on artificial intelligence. Specifically, it uses a Grok-based transformer model. This model learns relevance through engagement patterns. There is no manual tuning by employees. Instead, automation decides what content appears.
Because of this, X says operations remain simpler. Despite these claims, timing raises questions. European regulators recently fined X over transparency failures. At the same time, Grok faces scrutiny over harmful image generation.
Lawmakers and regulators continue to investigate. Some critics now call the open sourcing symbolic.
They argue it may distract from deeper issues. Still, X says openness builds trust. Musk previously claimed transparency leads to better recommendations. Whether users agree remains unclear.