Ant Lingbo CEO: Data Is Core to Robot Brain as Startup Kicks Off 1.5B Yuan Funding Round
On Sept. 11, 2026, Zhu Xing, CEO of Ant Lingbo, Ant Group's embodied AI subsidiary, said at the Bund Summit that the company has launched its first funding round targeting 1.5 billion yuan and plans a second round by the end of 2026.
Zhu said the funding is not driven by a lack of capital or computing power, but by the need to attract and retain senior talent and to strengthen Ant Lingbo's governance as an independent company. Ant Lingbo, founded in December 2024, focuses on building a universal brain for robots, providing closed-source foundation models and toolchains while collaborating with robot body manufacturers.
Zhu argued that data, not training technology, is the core barrier in embodied AI. OpenAI's GPT-6, released in September 2026, can improve R&D efficiency at certain stages, he said, but it does not solve the fundamental data problem. Ant Lingbo believes embodied intelligence must be trained on physical real-world data to build its own models, rather than being a simple extension of digital intelligence.
On data strategy, Zhu said the company values data quality distribution over sheer scale. Because existing off-the-shelf data often suffers from poor technical quality, duplication and low funnel rates, Ant Lingbo has relied on customized data collection rather than purchasing ready-made data. He acknowledged the industry is still early and requires patience, but predicted that from the second half of 2026 through 2028, some scenarios with lower task complexity could enter a phase of rapid growth in productivity applications.
