Meta, Google, OpenAI among Big Tech firms seeing top staff leaving to launch AI startups

A significant trend is emerging as top researchers from major tech companies like Meta and Google depart to establish their own AI startups, capturing the attention of investors eager to capitalize on the burgeoning potential of early-stage AI ventures. With massive financial backing, these new companies are securing hundreds of millions in funding within months of their inception. In a notable development, David Silver, a former researcher at Google DeepMind, recently announced a groundbreaking $1.1 billion seed funding round for his new startup, Ineffable Intelligence. Another ex-DeepMind employee, Tim Rocktäschel, is reportedly in the process of raising up to $1 billion for his own venture, Recursive Superintelligence. Meanwhile, AMI Labs, founded by Yann LeCun after his departure from Meta, secured a $1 billion investment in March and is focused on creating AI systems capable of learning from continuous real-world data. The trend of former employees from prestigious AI labs such as OpenAI, DeepMind, and Anthropic forming new companies is becoming increasingly common. These startups, including Periodic Labs, Ricursive Intelligence, and Humans&, have collectively raised substantial funds, reflecting a robust interest from investors in the capabilities and insights of these founders. Elise Stern, managing director at French venture capital firm Eurazeo, highlighted that the fierce competition among major AI labs has inadvertently created opportunities for smaller, agile firms. With many larger companies narrowing their focus to immediate goals, critical areas of research are being sidelined. Stern emphasized that this shift opens doors for innovative exploration in neglected fields. Investments in AI startups founded by distinguished researchers have surged, with VCs allocating $18.8 billion to such companies since early 2025, surpassing the previous year’s figures. Founders with deep experience in cutting-edge labs possess unique insights into scalable solutions and recognize untapped opportunities within their former organizations. However, the ambition for commercial success among leading AI labs poses challenges for researchers in those environments. Alexander Joël-Carbonell of HV Capital noted that the pressure to achieve rapid performance improvements leaves little room for exploratory research, particularly beyond the prevailing large language model (LLM) paradigm. For instance, Ricursive Intelligence, which raised $335 million in two rounds after its September founding, is developing AI tools for chip design. Co-founders Anna Goldie and Azalia Mirhoseini, both former employees of Anthropic and Google DeepMind, understand that neutrality is crucial for gaining trust among potential clients. Similarly, Periodic Labs is aiming to innovate in autonomous lab technology and has already secured $300 million in funding shortly after its launch. As the field of AI evolves, there is growing skepticism among researchers regarding the viability of merely scaling existing LLM approaches to achieve superior AI capabilities. AMI Labs, led by LaCun, believes the timing is ideal for advancements in AI, particularly as the technology seeks to transition from digital environments to real-world applications in robotics and healthcare. Ineffable Intelligence, focusing on reinforcement learning, aims to differentiate itself from traditional models that rely heavily on internet text data. This emerging landscape is indicative of a shifting paradigm in AI, where former tech giants are no longer the sole players in the game.