Igor Babuschkin, one of the founding minds behind Elon Musk’s artificial intelligence startup xAI, has stepped down to start a new chapter dedicated to AI safety. In a post on X, Babuschkin confirmed his departure, writing, “Today was my last day at xAI,” and announcing the launch of Babuschkin Ventures—an investment firm that will back research and startups working on the safe development of artificial intelligence.

The AI researcher, who previously worked at Google’s DeepMind and at OpenAI, played a pivotal role in shaping xAI from its earliest days. Reflecting on his journey, Babuschkin recalled the “early scramble” to build infrastructure and AI models, adding that he was responsible for many of the “foundational tools” used to launch and manage training jobs. Over time, he led engineering efforts across infrastructure, product development, and applied AI initiatives.

His exit marks another senior-level departure for xAI, which Musk launched in 2023 to challenge what he views as Big Tech’s overreach in the AI space. Musk has been an outspoken critic of the industry, often accusing major players of imposing excessive restrictions and neglecting safety measures.

This latest resignation follows that of Robert Keele, xAI’s head of legal, who stepped down earlier this month. Just weeks before, Linda Yaccarino also vacated her role as CEO of X—the social media platform Musk integrated into xAI last year. These leadership changes mirror ongoing transitions across Musk’s portfolio, including Tesla.

Responding to Babuschkin’s announcement, Musk kept his message short but notably warm: “Best of luck, Igor.” The cordial tone stood out to many observers, given Musk’s sometimes sharp public exchanges with former colleagues.

Babuschkin’s decision comes at a time when competition in AI is reaching fever pitch. Industry giants like OpenAI, Google, and Anthropic are pouring resources into advanced AI systems expected to shape the next generation of technology. xAI has been positioning itself as a challenger, promoting its chatbot Grok as a more open alternative to products like ChatGPT.

While Babuschkin has not provided detailed plans for his new firm, the mission of Babuschkin Ventures—to fund safe AI development—aligns closely with a growing global push for responsible AI practices. Governments, researchers, and advocacy groups are increasingly focused on ensuring that powerful AI models are developed with robust safety protocols to avoid potential misuse or unintended harm.

For Musk, Babuschkin’s departure represents another challenge in retaining top-tier talent at xAI, especially as rivals with far greater financial resources continue to lure experienced engineers and researchers. For Babuschkin, however, it’s a shift from building AI systems to enabling others to build them responsibly—a goal that could attract interest from both investors and policymakers who are seeking to influence the ethical trajectory of AI.

As the AI race accelerates, Babuschkin’s move highlights a crucial question facing the industry: not just how powerful AI can become, but how safely it can be built.