Why AI-Native Organizations Don't Always Need Flatter Teams

AI-native organizations are often considered to have flatter teams, fewer people, and fewer boundaries. However, this simplification can be misleading when it comes to understanding what AI can truly achieve and what organizational models should look like. The presence of boundaries in organizations is not always due to a lack of knowledge or context, but also because of diverse priorities, accountability requirements, and regulatory or physical constraints.

Some of these boundaries may fade away as AI becomes more prevalent, while others remain essential for maintaining order and preventing mistakes.

When deciding whether to flatten teams, it is crucial to consider the problem each boundary was designed to solve. AI might make it easier to cross certain boundaries, but that does not necessarily mean they should disappear. For example, a salesperson specializing in sales, accounting, law, security engineering, and supply-chain planning is not feasible due to limited attention, memory, training time, and available context.

Companies divide knowledge across roles to ensure that people deal with familiar problems themselves and call in specialists when deeper expertise is required.

AI can significantly improve execution by allowing a single agent to cover more work. For instance, a customer-service agent used to coordinate with multiple teams to handle a refund request. With AI, the agent can check the customer, order, payment, and policy and then trigger a refund, update the record, and create a ticket—all within one streamlined process.

However, granting an agent more authority does not automatically mean they should own the rules or policies involved. It is essential to maintain clear ownership and decision-making to avoid confusion and ensure accountability.

AI can streamline coordination by reducing the need for handoffs and improving communication between teams. However, this does not equate to consensus. AI is most effective when it is used to establish facts, apply rules, and execute policies. The real challenge arises when businesses have not reached a consensus on certain rules or when exceptions need to be handled.

In these situations, AI can assist by comparing options, but ultimately, the business must make decisions based on its priorities, such as revenue, margin, market share, customer satisfaction, or risk management.

It is also important to remember that process, decision rights, and organizational structure can change independently. While AI can streamline processes by removing handoffs and automating checks, it does not necessarily alter decision-making authority or the overall organizational structure. Companies need to carefully consider how these elements evolve separately, as treating them as interconnected can lead to unintended consequences and unpredictable outcomes.