Anthropic’s Aggressive Moves Prompt X Exec to Call for Its Ban

Recent AI‑community buzz centers on Anthropic after a tweet claimed the company banned Bun, a JavaScript runtime, from third‑party backends; the claim later turned out to be a satirical Onion‑style story.

The rumor felt plausible because Anthropic reportedly acquired Bun at the end of 2025 and plans to integrate it into Claude Code, and the company has recently taken aggressive actions against Elon Musk’s xAI.

First strike – cutting off xAI: According to tech media, xAI developers had been using Claude via the Cursor editor. Anthropic abruptly revoked xAI’s access. An internal Slack message from co‑founder Tony Wu warned the team, “Many of you have probably noticed that Claude isn’t responding in Cursor. Cursor says this is Anthropic’s new policy targeting all major competitors.” The move forced xAI to accelerate its own tooling, with Musk promising a major upgrade to Grok Code next month.

“It’s time to ban Anthropic on X.” – Nikita Bier, product lead at X (Twitter)

Second strike – suppressing open‑source: The popular GitHub project Crush (a terminal coding agent) received a pull request titled feat: remove claude code support. The maintainer explained, “Per Anthropic’s request, we removed Claude Code support to comply with its service terms.”

Similarly, the open‑source editor OpenCode began returning 400 errors when users tried to use Claude subscriptions, with the response:

{
  "message": "This credential is only authorized for use with Claude Code..."
}

A paying subscriber wrote an open letter to Anthropic, “As a subscriber I can’t understand why I can’t use my favourite client. If you persist, I’ll vote with my wallet.”

In contrast, OpenAI and GitHub embraced openness. OpenCode author Dax announced version 1.1.11, which now supports ChatGPT Plus/Pro subscriptions. GitHub lead Jared Palmer replied on Twitter, “Tell us how to implement ‘Login with Copilot’; we’ll officially support you.”

This contrast illustrates two strategic philosophies:

Anthropic’s strategy: A closed‑garden approach – the best experience is only through its official Claude Code CLI, regardless of usability.

OpenAI/GitHub’s strategy: An open‑ecosystem model – developers can use any tools they prefer as long as they consume the service.