XAI co-founder leaving to start AI safety VC firm

Igor Babuschkin, co-founder of xAI, announced on August 13, 2025 that he is leaving the company to start a VC firm focused on AI safety.

XAI, started in 2023 by Babuschkin and Elon Musk, has developed AI models and the chatbot Grok, which has faced recent controversies over its responses and features.

Babuschkin said his new firm, Babuschkin Ventures, will support AI safety research and invest in startups working on advanced technologies.

Before xAI, Babuschkin worked at Google DeepMind on AlphaStar and at OpenAI.

🔗 Source: TechCrunch

🧠 Food for thought

1️⃣ High-burn AI companies face mounting leadership retention challenges

xAI’s struggle to retain key technical talent reflects broader pressures facing rapidly scaling AI companies.

Babuschkin managed a team of 29 engineers and was instrumental in building xAI’s Memphis supercluster and Grok chatbot, making his departure particularly impactful for the company’s technical capabilities 1.

His exit comes amid what sources describe as “a wave of executive departures” at xAI, suggesting systemic retention issues rather than an isolated incident 2.

This leadership instability occurs as xAI burns through $13 billion annually while maintaining a $200 billion valuation, creating intense pressure to deliver results that can strain even experienced technical leaders 1.

The situation illustrates how AI companies’ aggressive growth timelines and massive capital requirements can create challenging working conditions, even for founders who helped build the organization from the ground up.

2️⃣ AI safety emerges as a distinct venture capital category

Babuschkin’s transition from building AI systems to funding AI safety research signals the maturation of safety concerns into a legitimate investment thesis.

His new firm, Babuschkin Ventures, will specifically target startups focused on “AI safety research” and ethical frameworks, representing a shift from his previous role developing cutting-edge AI capabilities 3.

This career pivot reflects growing industry recognition that safety considerations require dedicated capital and expertise, separate from traditional AI development funding.