Keynote | Human AI task allocation: deciding what stays human when everything can be automated Every support leader now has the technical option to automate almost anything. Very few have a clear, defensible answer for what should stay human, and that gap is where most AI support rollouts either earn trust or lose it. This session looks at how leading teams are drawing that line in practice, which task categories are safe to hand fully to AI, which need a human in the loop by design rather than by exception, and how that allocation should shift as the technology and the team's confidence in it both mature. Key takeaways: - A practical framework for sorting support tasks into fully automated, human in the loop, and fully human categories - Why the right allocation today is not the right allocation in twelve months, and how to plan for that - What signals tell you a task has been automated too early - How to communicate the allocation decision internally so it does not read as a headcount story