A rollout can have a budget, a launch date, and a training plan while leaving decisions about workloads, responsibilities, and quality unresolved. Employees have to integrate the technology while meeting existing delivery expectations, and managers inherit responsibility for working practices they have not helped define. I consider readiness part of the investment decision, including whether the organization has the time, management capability, and clarity of responsibility to make adoption work.

General-purpose AI may possess extraordinary general knowledge, but it doesn’t inherently understand your organization’s definitions, exceptions, informal practices, customers, or decision rules. Much of that context lives in employees’ heads, scattered documents, and assumptions nobody has needed to explain. Making it usable requires organizational judgment: identifying what matters, resolving conflicting information, and documenting how decisions are made. The people who understand the business need an active role in shaping the context their AI works with and evaluating whether its responses are useful.

Usage reports leave out much of the work needed to make AI output usable. A fast draft may require substantial correction by a more experienced colleague, so apparent time savings can conceal a transfer of effort. Managers need enough understanding of the technology and the task to recognize that transfer and assess the employee’s judgment. My standard for value includes the quality of the finished work and the total effort required to produce, verify, and use it.

An employee under pressure to deliver may find a useful tool before the organization has provided an approved option. Others may be using approved tools in ways nobody has examined. The pressures behind those choices matter alongside the risks they create. Oversight needs to make responsible use achievable through clear boundaries, suitable access, practical support, and enough trust for employees to disclose their methods and mistakes. An organization cannot make informed adoption decisions when it has little visibility into how the work is being done.

A process can appear efficient because experienced employees absorb its gaps. They chase missing information, reinterpret requests, negotiate exceptions, and remember details the system never records. Delivery continues, concealing how much effort and individual knowledge it requires. Those workarounds reveal both necessary professional judgment and weaknesses in the formal process. Any redesign needs to account for that hidden work, preserve useful discretion, and address the gaps before attempting to standardize or automate them.

A manager can understand delegation and still take work back because they remain answerable for every mistake and have little time to coach. When delivery repeatedly depends on that intervention, the organization reinforces the habit it wants the manager to change. Leadership development needs to address those pressures, including the authority, support, and consequences surrounding a decision. Practice and feedback become useful when managers can apply them to the conversations and responsibilities already on their desks.

Employees interpret values through the decisions they see supported, rewarded, or reversed. An organization may encourage ownership while requiring approval for routine decisions, or ask people to speak openly while responding poorly to unwelcome information. Different managers can reinforce different versions of the same value. I believe values need to be examined where they become difficult to uphold: under deadline pressure, across departmental boundaries, or when someone challenges a senior person’s judgment. Those situations reveal what the organization actually supports.

Teams can support a strategy while remaining accountable for targets, budgets, and workloads established before it existed. Each department then makes reasonable local decisions that collectively undermine the intended change. Agreement leaves important trade-offs unresolved, including which priorities take precedence and who has authority when they collide. A strategy needs corresponding decisions about resources, responsibilities, and what work will stop. Otherwise, employees are left to negotiate competing expectations that require leadership decisions.

A training request can carry an untested assumption that people are underperforming because they lack knowledge. Conflicting instructions, inaccessible information, inadequate feedback, and unworkable workloads can produce similar symptoms. Even where capability is the issue, people may need practice with difficult decisions and feedback on their reasoning. I expect learning to begin with a defined business problem and a clear account of what people need to do differently, with the surrounding management and operating conditions examined alongside it.

Customers experience the combined effect of decisions made across the business, even when each department is meeting its own targets. Sales may promise flexibility, operations may prioritize consistency, and service staff may lack authority to resolve the resulting tension. Experienced employees compensate, making the experience depend on who happens to be involved. A customer promise needs compatible departmental priorities, reliable handoffs, and enough employee judgment and authority to handle situations that a script cannot anticipate.