X Upgrades Platform Features With Grok AI for Translations and Photo Editing

X's head of product Nikita Bier announced on April 7 that the platform was rolling out Grok-powered automatic translation worldwide and launching a new photo editor for iOS that lets users rewrite images through plain-language prompts. Both features deepen xAI's grip on X's core infrastructure, replacing Google Translate as the platform's default translation engine and extending Grok's generative reach to every image posted through the composer.

The translation rollout is global and on by default. Users who want to read posts in their original language must manually opt out, language by language, through a gear icon on each translated post. "We're rolling out auto-translate worldwide to give posts in any language global reach on X," Bier wrote. "The translations are powered by Grok and have improved substantially over the last couple months." X has not published error rates for Grok's translations compared with the Google Translate system it replaced, nor has it specified how translated posts will be labeled to distinguish machine output from original text.

The photo editor, accessible through a paintbrush icon in the post composer, includes drawing tools, text overlays, a blur function for redacting faces or documents such as credit cards and Social Security numbers, and a Grok-powered natural language editing mode. A demo video shared by Bier showed an image reframed to appear as though it were hanging in a museum. The tool currently ships on iOS only.

The rollout arrives against a backdrop of unresolved trust problems with Grok. In 2024, the AI spread false election information for ten days before being corrected, prompting five state secretaries of state to write directly to the platform. xAI also faces a class action lawsuit from three teenagers who allege their photographs were used to generate child sexual abuse material through Grok before X restricted the feature to paying subscribers. The platform has offered no independent audit mechanism for Grok's outputs on either translation accuracy or image manipulation.

What makes the timing especially sharp is Bier's own words from just five weeks ago. Announcing in March that creators who post AI-generated war footage without disclosure would face a 90-day suspension from X's Creator Revenue Sharing program, he wrote: "During times of war, it is critical that people have access to authentic information on the ground. With today's AI technologies, it is trivial to create content that can mislead people." That policy targets third-party creators; it does not address what happens when X's own Grok silently alters a translated post's tone or generates a manipulated image before it is published.

The deeper structural concern is one of consolidation. xAI acquired X in an all-stock deal valued at $80 billion, meaning every natural language prompt typed into the new photo editor feeds directly back into xAI's training pipeline. X has systematically replaced third-party tools, from search to post composition to fact-checking, with Grok, positioning the model not as an optional assistant but as the platform's connective tissue. With midterm elections six months away and no public standard for labeling Grok-translated or Grok-edited content, the question of what counts as authentic on X has no clean answer.