In a significant move towards transparency, social media platform X, formerly known as Twitter, will publicly release the code for its core recommendation algorithm within the next seven days. The announcement was made by the platform's owner, Elon Musk, in a post on Saturday.
Weekly Updates and AI-Driven Future
Musk stated that this disclosure will not be a one-time event. The plan is to repeat the process every four weeks, accompanied by detailed developer notes to help the public track changes. This initiative appears to be a step towards fulfilling Musk's long-standing promise to make X's algorithms open source, though previous follow-through has been inconsistent.
The billionaire's vision for X's content feed is increasingly powered by artificial intelligence. He confirmed that the company is working to incorporate more AI, specifically using Grok, his proprietary AI chatbot, into the recommendation engine. The goal, as Musk wrote in September, is for the recommendation system to "be purely AI."
Addressing Bugs and Regulatory Pressure
The decision to open the algorithm comes at a time when X has faced user complaints and regulatory challenges. Some users had reported seeing fewer posts from accounts they follow. In October, Musk acknowledged a "significant bug" in the "For You" algorithm and promised a fix.
More critically, X and Musk have repeatedly clashed with global regulators over content moderation, transparency, and the spread of misinformation. European regulators have intensified their scrutiny. In a notable instance from July, French authorities requested X share its algorithm as part of an investigation into alleged bias and manipulation, a request the company refused, calling it politically motivated.
Grok's Role and Growing Controversies
Musk's October posts elaborated on Grok's expanding role. He indicated that improvements in user feeds are increasingly due to AI tools like Grok rather than manual adjustments. The ambitious plan involves having Grok evaluate all 100+ million daily posts on X to curate a highly personalized feed for each user, a feature Musk claimed would "profoundly improve the quality" and was targeted for a November rollout.
However, Grok itself is under fire. Its image-generation feature has been criticized for contributing to a flood of AI-generated sexualized images, including depictions of women and children, on the platform. This led Indonesia to block access to Grok entirely. Just this week, UK Prime Minister Keir Starmer urged X to "get their act together," and Technology Secretary Liz Kendall warned the UK government could block services that do not comply with local laws. In response, as of last Friday, X restricted image generation and editing on Grok to paid subscribers only.
A Step Towards Openness or Necessity?
While Musk framed the algorithm release as a push for transparency, he did not specify the exact reasons behind the timing. The move could be seen as a preemptive step to address mounting regulatory pressure in key markets like Europe and the UK, where demands for algorithmic accountability are rising. By making the code public, X may aim to demonstrate there is no hidden bias, though its effectiveness in appeasing regulators remains to be seen. The world will be watching closely when the first lines of code are revealed in one week.
