Why large organisations reject new ways of working
Large companies are very good at two things: doing the same thing reliably, and doing it at the lowest cost. Both are strengths. Both make it hard to do something new. A new way of working is unpredictable at first and inefficient at first, which is exactly what a well-run organisation is designed to eliminate.
How new projects are rejected
It rarely looks like opposition. It looks like reasonable management. A strategic alignment review that takes three months. A budget moved to a more urgent priority. A requirement to integrate with every existing system before going live. A steering committee whose members each have something to lose. Each step is defensible. Together they ensure the project never reaches the point where it could prove anything.
Not every failed AI project was killed by the organisation. Some failed because the technology was not ready, the numbers did not work, or the wrong process was chosen. But when the technology works, the numbers work and the project still stalls, internal resistance is usually the reason.
Build outside, then bring it in
The way through is structural, not motivational. Put the new work outside the normal reporting lines, in a small team that answers directly to the chief executive. Fund it from the top, never from the budget of the division it may replace. Give the team a share in the result, not just a salary. And protect it until it has numbers to show.
The existing organisation keeps running and keeps earning while this happens. Nobody's position is threatened by a slide deck. When the new process clearly works better, it takes over, and the argument is made by results rather than by memos.
What the sponsor must do
Most of these efforts die when the sponsor loses interest or leaves. The defence is speed: get to results the board can see, measured in outcomes, not in milestones.