Autoscience Secures US$14 Million to Launch Fully Autonomous Artificial Intelligence Laboratory
San Mateo, Saturday 21 March 2026
On 18 March 2026, Autoscience secured US$14 million to launch a fully autonomous laboratory where artificial intelligence replaces human researchers, compressing decades of discovery into mere months.
Redefining Research with Autonomous Architecture
The recent capital injection into the San Mateo-based startup was led by General Catalyst, alongside participation from Toyota Ventures, the Perplexity Fund, MaC Ventures, and S32 [1][2][3][4]. The enterprise is constructing a dual-layer virtual laboratory designed to eliminate human bottlenecks in artificial intelligence research [4]. As explained by Co-Founder and Chief Executive Officer Eliot Cowan, the platform operates via an ‘AI Scientist’ that analyses literature and generates hypotheses, paired with an ‘AI Engineer’ that translates these concepts into production-ready machine learning models [4]. This architecture is engineered to condense timelines that traditionally span years into a matter of months [4].
The Broader Shift Towards Agentic Workflows
The investment in Autoscience reflects a wider venture capital trend favouring autonomous agentic frameworks over passive software-as-a-service (SaaS) tools [GPT]. On 19 March 2026, Edra secured US$30 million in a Series A round led by Sequoia to build artificial intelligence agents capable of automating complex business processes [8]. Simultaneously, the infrastructure required to monitor these autonomous systems is attracting capital; Respan raised US$5 million to build proactive observability for artificial intelligence systems, whilst Laminar secured US$3 million for open-source agent observability, representing a combined investment of US$8 million in the observability sub-sector [7].