Microsoft and NYU tested what happens when AI becomes employee #1
The research shows we're moving from AI-as-tool to AI-as-colleague, which means rethinking how we structure accountability and human oversight.
Harvard Business Review covered a Microsoft-NYU collaboration where 30 MBA students built startups using AI as their first team member. The students worked in six teams, each given access to Microsoft 365 Copilot and tasked with building companies from scratch—solar installation, personal finance apps, wellness wearables.
What emerged shifts the conversation about early team structures. Students found AI became their "first hire," handling analyst and strategy work that traditionally requires human expertise. One team let AI design their org chart based on resume analysis. Others used it for financial modeling and brand development. The result: founders could move from idea to launch faster with smaller human teams.
The implications for product and legal teams run deeper than efficiency gains. Students had to develop new skills in evaluation and oversight rather than execution. As one noted, "You have to be the expert. You have to be the one to say, 'this is right, this is wrong.'" That shift—from AI as tool to AI as colleague—surfaces questions about how we structure accountability and human oversight when machines take on work we used to assign to people.
J.P. Eggers
