Skyfall AI to Test AI CEO Model with $1M Acquisition

Huo Xing Cai Jing reports that on July 20, Skyfall AI, a startup founded by former members of Microsoft’s AI team, announced plans to test the feasibility of an AI-run company by acquiring a small B2B SaaS or e-commerce business, with the goal of having AI assume responsibilities equivalent to those of a CEO. Skyfall intends to invest up to $1 million to acquire a real business and entrust its AI system with core operational functions—including pricing, marketing, customer support, finance, and operations—while minimizing human intervention. The company aims to double the revenue of the acquired business, with the entire experiment being publicly documented. Sam Pasupalak, co-founder and CEO of Skyfall, stated that while the industry currently focuses on AI replacing individual roles, true autonomous enterprises require AI capable of end-to-end management. He believes only by deploying AI in real-world commercial environments can the viability of “autonomous companies” be validated. The company notes that current large language models (LLMs) face limitations in complex enterprise settings: although they can handle known knowledge and fixed tasks, they lack the capacity for continuous learning in response to competitive shifts, changing consumer behavior, and dynamic market conditions. To address this, Skyfall proposes the concept of “Enterprise World Models,” aiming to build AI systems capable of understanding business operations, predicting the impact of decisions, and engaging in long-term planning. Skyfall was founded by the original team behind Maluuba, which Microsoft acquired for approximately $160 million in 2017 and integrated into its Canadian AI research division. While major AI players including OpenAI, Anthropic, and Google are advancing AI employees and enterprise automation, Skyfall believes the future competitive frontier lies not in AI assisting human workers, but in building enterprises that can operate autonomously. The company emphasizes that AI is not intended to fully replace human leaders, but rather to reduce the burden of daily operations so humans can focus on strategy, oversight, and accountability. This experiment will test whether AI can effectively manage complex organizations over the long term in real-world commercial environments.