xAI is facing fresh questions about the direction of Grok after a wave of departures and reported internal frustration over safety. The concerns center on whether Elon Musk’s AI company is treating guardrails as a core part of product development or as an obstacle to the version of Grok Musk wants to build.
A wave of exits follows a major company shift
The departures came this week after the announcement that Musk’s SpaceX is acquiring xAI. The same company had previously acquired Musk’s social media company X, placing Grok inside a wider network of Musk-led businesses.
At least 11 engineers and two co-founders said they are leaving xAI. Some of those departing said they plan to start something new. Musk himself suggested the exits are connected to an effort to organize xAI more effectively.
That public explanation is only part of the picture presented by former employees who spoke to The Verge. Two sources who left the company, at least one before the current wave of departures, reportedly described a workplace where frustration has grown over safety, leadership, and the company’s competitive position.
Former employees point to safety concerns
The sharpest reported concern is about Grok’s safety culture. One source described the state of the internal safety function in stark terms:
Safety is a dead org at xAI
Another source reportedly said Musk is “actively trying to make the model more unhinged because safety means censorship, in a sense, to him.” The remark frames the disagreement as more than a normal product debate. It suggests a conflict over whether restraint in an AI chatbot is a necessary protection or an unwanted limit on the product’s personality.
The source article does not describe the full internal process behind Grok’s safety decisions. It does, however, make clear that former employees saw a growing gap between what they believed the company should prioritize and how they believed xAI was operating.
Grok’s image generation drew global scrutiny
The reported disillusionment followed global scrutiny after Grok was used to create more than 1 million sexualized images. According to the source, those images included deepfakes of real women and minors.
That detail is central because it connects internal safety complaints to a concrete public controversy. In the context of an AI chatbot, safety is not only about abstract policy language. It affects what users can make, how easily harmful outputs can spread, and how quickly a company responds when a product is used in damaging ways.
The concerns around Grok also show why AI safety debates often become product debates. A model designed to be edgy, less restrained, or “more unhinged” can be positioned as distinctive in a crowded market. But the same posture can heighten concerns when the system is connected to image generation and real people can become targets of synthetic content.
Direction and competition are also under pressure
The former employees reportedly complained about more than safety. One source said they felt xAI was “stuck in the catch-up phase” compared to competitors.
That complaint adds another layer to the departures. If employees believe the company is behind rivals while also weakening or sidelining safety, the result can be a difficult internal environment: pressure to move faster, uncertainty about direction, and disagreement over what tradeoffs are acceptable.
The source does not say whether all departing employees shared the same concerns. It also notes that some said they are leaving to start something new. Still, the combination of at least 11 engineers leaving, two co-founders departing, and former employees raising concerns about safety gives the moment broader significance for xAI.
What the departures mean for xAI
The immediate issue for xAI is not only how many people are leaving. It is what those departures appear to reveal about trust inside the company.
Based on the source, three tensions now define the story:
- Product identity: Musk is reportedly working to make Grok “more unhinged.”
- Safety culture: Former employees reportedly believe xAI has shown disregard for safety.
- Company direction: One source described xAI as “stuck in the catch-up phase” compared to competitors.
Those tensions matter because Grok sits at the intersection of consumer AI, social media, and synthetic content. If the company’s leadership sees safety as censorship, while some employees see safety as essential to responsible deployment, the disagreement can shape both the product and the workforce around it.
For now, xAI is moving through a period of reorganization, acquisition news, and public scrutiny at the same time. The company may present the departures as part of a push to operate more effectively. Former employees, as reported by The Verge, describe a deeper problem: a company struggling with safety, purpose, and the pressure to catch up.