The recent wave of departures at xAI, Elon Musk’s AI startup, has raised concerns about the company’s internal culture, strategic direction, and leadership. Over the past few days, multiple cofounders and employees have announced their exits, leaving xAI with only six of its original 12 cofounders still on staff. Cofounders Yuhuai (Tony) Wu and Jimmy Ba both cited personal reasons, with Wu stating it was “time for [his] next chapter” and Ba calling for a “recalibration” of his focus. Their departures come amid broader staff attrition, with several former employees announcing they are launching their own AI ventures. The exodus follows the recent announcement of a merger between xAI and SpaceX, which has reportedly given xAI shareholders $250 billion in new shares. This windfall may have enabled employees to pursue independent projects, as some have openly stated their intent to build new AI companies. Vahid Kazemi, a former employee, wrote on X that all major AI labs are building “the exact same thing,” calling it “boring,” and that he’s starting something new to inject more creativity. Another former staffer said he’s leaving to “build something new, focused on accelerating science.” A group of ex-employees is also launching Nuraline, an AI infrastructure company, with one noting that while xAI showed promise in solving measurable problems, it failed to evolve beyond model weights, ignoring the broader system-level intelligence needed for true advancement. Internal dynamics appear to have contributed significantly to the unrest. A former employee who left earlier this year described a culture of disillusionment, citing xAI’s focus on NSFW content in Grok as a symptom of deeper issues. He claimed the safety team had been dismantled, leaving only basic filters for content like CSAM, and said “safety is a dead org at xAI.” He also criticized the lack of innovation, saying the company was stuck in a “catch-up phase” with no breakthroughs compared to rivals like OpenAI or Anthropic. Another former employee echoed these concerns, stating that xAI was not taking bold risks and that decisions were often made in an all-company group chat on X, with Musk as the final arbiter. This decentralized, fast-moving approach, while enabling rapid iteration, reportedly led to infighting and stalled progress. He also noted the absence of human review in the development process and the pressure to “shut up and do what Elon wants.” Musk’s recent vision for the merged entity includes ambitious plans such as space-based AI data centers and a Moon-based AI factory and city. He shared a recording of an internal all-hands meeting where he outlined four new xAI divisions: Grok Main and Voice, Coding, Imagine (image and video), and Macrohard, which aims to digitally emulate entire companies. However, the absence of any mention of a safety team in the new org chart reinforces concerns about oversight. Critics argue that Musk’s leadership style and focus on rapid deployment over safety and originality have alienated top talent. As one source put it, “Trying to do what OpenAI was doing a year ago is not how you beat OpenAI.” With a growing number of former employees striking out on their own, the exodus may signal not just a reorganization, but a fundamental shift in the future of xAI and its role in the broader AI landscape.