Yes, traditional models do risk disincentivising it, but again change here is difficult.

If partner reward and promotion still lean heavily on billed hours, personal utilisation and short-term origination, then you are effectively telling people not to automate too much, not to share efficiency too widely, and not to redesign work in ways that reduce time. That becomes a real problem once firms are trying to deploy AI agents that can remove or compress repeatable junior-heavy tasks. Several of the firm diagnostics we’ve reviewed make exactly that point… AI is pushing firms towards value-based pricing, different leverage models and more explicit links between pricing, profitability and partner behaviour.

The firms thinking seriously about this are starting to shift incentives away from pure hours and towards a broader mix of measures. That usually means rewarding profitable revenue, pricing discipline, matter margin, client outcomes, adoption of approved AI workflows, and contribution to building reusable capability across the firm. In other words, if someone helps create an AI-enabled way of delivering work faster, more consistently and at better margin, the system needs to recognise that as value creation, not treat it as lost time.

I do not think this means firms suddenly abandon current models or start paying bonuses for every prompt. But it does mean the old model needs updating. The most sensible approach is usually evolutionary – keep the parts of compensation that protect collaboration and firm culture, but add clearer weight for commercial performance, innovation, adoption, and client impact. Some firms are already experimenting with this more visibly, including linking part of reward to AI adoption and practical use of new tools.