top of page

Re-imagining governance in the age of AI: Balancing speed, control and learning


Cartoon that says the better the breaks, the faster you go with a sled going down a slope

An organisational priority to improve delivery


In today’s fast-paced world, accelerated even further by AI, governance is often viewed as a barrier to achieving fast flow - a function stuck in bureaucracy and control. But when we look at governance through a different lens, it becomes clear that governance is not the roadblock.


It’s the structure that enables organisations to thrive in an environment where we need to balance both speed and control. As AI increases the speed at which we can analyse, experiment, make decisions and build, good governance becomes more important, not less. 


By reframing governance as a leadership tool, rather than just a compliance function, we unlock the ability to align decision-making, create flow, and ultimately deliver value at scale.


The hidden cost of overlooking governance


A recurring theme in many large transformations is the delayed re-imagining of governance. While teams are restructured, delivery practices evolve, and ways of working become more adaptive, the underlying governance structures - those that control funding, risk, and decision-making - often remain stuck in the past.


This disjointed approach leads to a system that is both new and old, weighed down by outdated and heavy assurance models. This holds back progress, drains time, money, and, perhaps most significantly, people’s energy.


AI will make this governance debt even more visible. Slow decision making, heavy assurance processes and a lack of empowerment become increasingly obvious constraints. There is little value in accelerating delivery if work then sits waiting for a decision, approval or control.


Speed, control and learning: Partners, not opposites


When governance is done right it doesn’t stifle delivery, it accelerates it. The key to success here is clarity. Clear governance provides the guardrails needed for innovation to flourish safely. It helps everyone understand how decisions are made, when to escalate issues, and how value is measured and assured. 


This clarity is especially crucial when teams are operating with more autonomy, as it ensures that everyone is aligned on the "what" and the "why."


In an AI-enabled organisation, I would add learning to this equation. Governance needs to help organisations move quickly and safely, but it also needs to help us learn. Are we still doing the right things? What are we learning from our experiments? Are new risks emerging? Do we need to stop, pivot or continue?


The faster the pace of change, the faster these feedback loops need to become.


Shifting from control to enablement


Modern leadership in governance isn’t about controlling every aspect of the process - it’s about enablement. Leaders don't need to be in every detail. Instead they need to create the conditions for flow. 


This means leaders in governance need to create space for alignment and decision-making closer to where the work happens, rather than enforcing micromanagement. Decision rights, thresholds and guardrails need to be clear enough that teams can act without unnecessarily escalating decisions through the hierarchy.


This shift in mindset and language moves us away from command-and-control towards a more collaborative approach that maintains standards while enabling teams to deliver their best work.


Importantly, governance is not something that happens to teams. Risk, controls and assurance need to increasingly become part of the flow of work and everyone's responsibility - rather than another process or checkpoint at the end.


Trust, data, transparency and sensing


For leaders to embrace this shift, trust is key, and trust is built on visibility. Data plays a critical role in this, but many organisations still have rich delivery data sitting across their tooling ecosystem without using that data to inform governance decisions.

The opportunity now goes beyond dashboards and reporting. When governance is connected to the data generated through the flow of work, organisations can move from point-in-time reporting towards continuous sensing.


AI creates the opportunity to identify emerging risks, dependencies, changes in delivery health and deviations from strategic outcomes earlier. This means governance can increasingly shift from reactive to responsive to proactive.


The goal should be fewer governance papers explaining what happened last month, and more real-time insight helping us decide what to do next. This creates faster feedback loops, strengthens trust between teams and leaders, and enables more effective decision making.


AI can increasingly support this by sensing risks, surfacing insights and informing decisions. But accountability cannot be outsourced to an AI agent. Leaders still need clarity on who owns the decision, what guardrails apply and where human judgement is required.


Designing governance to support flow


Effective governance doesn’t just protect flow, it enables it E2E. It’s about:


  • Aligning investment with outcomes, and continuously testing whether capacity and investment remain aligned to the highest value outcomes

  • Pushing decisions closer to where the context is, establishing clear decision rights, thresholds and guardrails so teams can act without unnecessary escalation

  • Embedding controls into the flow of work by automating and integrating assurance wherever possible rather than adding rigid checkpoints afterwards

  • Creating continuous transparency and sensing, using connected data and AI to surface risk, flow, value and emerging issues without creating another reporting burden.


Good governance ensures organisations do the right things, at the right time, in the right way. In the age of AI, this matters even more. As the pace of delivery accelerates, governance needs to keep pace with the organisation we are trying to enable.


The opportunity isn't to remove governance. The opportunity is to re-imagine it as an enabler of speed, control and learning.


Further learning



 
 
bottom of page