Governance Evolution

Rules that cannot change become obstacles to the thing they were meant to help.

It is a familiar story. An organisation decides it needs better governance, and so it writes a policy. The policy is thoughtful, comprehensive, and carefully negotiated. It sits in place for a year, then three, then ten. During those years the world around the policy changes. New technologies arrive. New risks appear. What used to be the right answer becomes wrong, and what used to be banned becomes standard practice somewhere else, and the policy slowly calcifies into a reason people cannot do things rather than a guide to doing them well. At some point, the people inside the organisation begin to see the governance as the problem rather than the solution, and they start routing around it. The rules remain intact. Their effect is nil.

This is particularly dangerous with AI in education, because the technology is moving at a speed that no paper policy can match. Governance that can only be updated by rewriting the system is not really governing. It is blocking until somebody decides to ignore it. The alternative is to build governance that expects to change, with explicit hooks for updating rules as the environment shifts, without redesigning the whole system each time. This is harder in the short term and cheaper in the long term, and it almost never gets funded properly because it has no obvious deliverable. Sound governance is not the heavy book that sits on the shelf. It is the living thing that keeps adapting alongside what it is supposed to watch over. When you last had to work around a rule to get something useful done, was the rule wrong or just outdated?

Last week we said small living rules beat big frozen ones. This week we look at what happens to rules that cannot keep changing, and at the design of rules that can.