Companies hoping to build their business around artificial intelligence should start by stripping out management layers, according to Rivian chief executive RJ Scaringe, who told the Inc. 5000 Conference and Gala that a flatter organisational structure suits the AI era better than a traditional hierarchy.
Scaringe's argument is that AI changes the speed at which information moves through a company, and that a structure with fewer rungs between the top and the front line allows decisions to be made and acted on more quickly. In his view, the organisational change matters as much as the technology itself: a business can adopt the latest tools, but if its people are buried under layers of approval, the gains are slow to arrive.
The intervention lands as founders and executives across Britain and beyond weigh how to reorganise their workforces around AI. Many have focused on buying software or hiring specialists, but Scaringe's advice points to something cheaper and harder: redesigning who reports to whom, and giving teams closer to the customer more room to act.
Rivian, the electric vehicle maker Scaringe leads, is itself a capital-intensive manufacturer where engineering and production decisions have to be coordinated across large teams. His suggestion that flatter structures work better is therefore notable, because it comes from a sector where hierarchy and process are often seen as essential to safety and quality. The implication is that even complex industrial businesses can benefit from fewer layers when AI is doing more of the routine coordination.
The remarks also reflect a wider debate about how AI should be governed inside companies. As agents and automated tools take on more tasks, executives are being pressed to explain who is responsible when something goes wrong. Atlassian cofounder Mike Cannon-Brookes has argued that AI agents need clear ownership, telling Fortune that accountability must sit with a named human rather than with the tool itself. His company has rolled out a system that gives AI agents visible profiles and owner attribution so that edits can be traced in the same way as a colleague's.
That concern connects to Scaringe's organisational point. A flatter company may move faster, but it also concentrates responsibility in fewer places, making it more important that each team knows who owns which decisions and which automated processes. Founders who flatten their structures without clarifying accountability risk creating confusion rather than speed.
The appetite for AI-driven business building is growing. Idealab founder Bill Gross has said he believes AI can help mass-produce new startups and jobs at a scale previously impossible, a claim that underlines how much expectation is being placed on the technology to reshape company formation. Against that backdrop, Scaringe's advice is a reminder that the organisational side of the AI transition is not automatic.
For British business leaders, the practical question is how far to go. Flatter structures can improve responsiveness, but they also demand clearer roles and stronger communication. Scaringe's message to the Inc. 5000 audience was that companies wanting to be built around AI should look at their own hierarchies first, before assuming that new software alone will deliver the change they want.