Allbirds CEO launches AI startup with no employees: the future of work?

What happened?

Joey Zwillinger, co-founder and CEO of Allbirds, has launched an AI startup that, according to TechCrunch, has no human employees. The company operates exclusively with autonomous AI agents, from development to customer service. The startup has completed a seed funding round, the amount of which has not been disclosed, but is described as 'very large' (TechCrunch). Zwillinger acts as the sole founder, and the business plan is based on AI performing all operational functions. This move is unprecedented: no known startup had completely eliminated the human workforce from its inception. The announcement comes in a context where automation has advanced rapidly, but always with human oversight.

Why is it important?

This case represents an extreme experiment in total automation of a company. While many startups use AI to optimize processes (e.g., using chatbots for customer service or algorithms in logistics), none had completely eliminated the human workforce. Zwillinger's move could set a dangerous or innovative precedent, depending on perspective. Additionally, it raises legal and regulatory questions about corporate responsibility, labor rights, and the long-term viability of a model without employees. Historically, automation has replaced tasks, not entire companies. The Industrial Revolution eliminated trades but created others; the digital age did the same. However, this case goes further: the startup not only automates processes but lacks any human on its payroll, which could challenge labor laws that require a responsible employer. In countries like Spain or France, where labor law protects workers, such a model could be illegal. In the United States, the situation is more ambiguous, but responsibility for automated decisions (e.g., loans, contracts) falls on the company, which in this case has no employees to attribute errors to.

Potential consequences

• Productivity and scalability: Without labor costs, the startup could scale quickly, but the lack of human oversight could lead to systemic errors. For example, an AI agent in charge of development could introduce security vulnerabilities without anyone detecting them. Additionally, AI infrastructure, energy, and maintenance are significant costs that could outweigh salary savings.
• Ethics and employment: If the model succeeds, it could accelerate the replacement of workers in tech sectors, increasing structural unemployment. According to a McKinsey Global Institute study (2023), up to 800 million jobs could be automated by 2030, but most experts believe automation complements, not replaces, humans. This experiment could change that perception.
• Regulation: Governments could intervene to require a minimum number of human employees or clear responsibilities for automated decisions. The European Union is already working on the AI Act, which classifies AI systems by risk. A company without employees making autonomous decisions would likely fall into the high-risk category, requiring mandatory human oversight. In the US, the FTC could investigate deceptive practices if the startup does not disclose its automated nature to customers.
• Public perception: Brands associated with this model could face boycotts or criticism for social insensitivity. A Pew Research study (2022) indicates that 72% of Americans are worried about a future where robots replace humans in many jobs. Zwillinger's startup could become a symbol of that concern.

What readers should know

The case is still in its early stages; there are no concrete operational results. The lack of employees does not imply absence of costs: AI infrastructure, energy, and maintenance are significant. Additionally, the startup will likely face challenges in areas like innovation, creativity, and handling exceptions, where human intelligence remains superior. Investors have bet on the promise of efficiency, but the risk of failure is high. Compared to other total automation experiments, such as the e-commerce startup 'Brandless' (which tried to eliminate human teams but failed), this attempt is more radical. Even giants like Tesla have faced problems in their automated factories, where human intervention is still necessary. Legally, the startup could face lawsuits if the AI makes errors causing harm, as there is no human employee to blame. Responsibility would fall on the company, but without employees, who answers? Founders could be personally liable. Finally, the debate on the future of work intensifies: are we witnessing the birth of the human-less company or an experiment that will border on illegality? Only time will tell.

“What Zwillinger is doing is a radical bet. If it works, it could redefine the concept of a startup; if it fails, it will be a warning about the limits of automation.” — TechCrunch analyst

The case also highlights the need to rethink employment and social protection policies in the AI era. If this model becomes widespread, we could see increased inequality, where owners of capital (AI and algorithms) accumulate wealth without generating employment. On the other hand, it could free humans from tedious jobs, allowing them to engage in creative activities. However, for now, Zwillinger's startup is just an experiment. The coming months will be crucial to see if it can operate effectively without human intervention. The tech community watches closely, and regulators too.