WHAT RESILIENT BUSINESSES DO DIFFERENTLY


Businesses that move past this stage tend to share a few common traits. They recognise early that AI will change faster than traditional systems, and so instead of pursuing stability through rigidity, they design deliberately for evolution. Business logic is separated from model logic and assumptions are recorded rather than hidden.


They treat validation and observability as core disciplines, not simply to satisfy regulators, but to sustain trust. They know where their numbers originate, how they are produced, and when they deserve scrutiny.


WHERE PROGRESS REALLY STALLS


When businesses find themselves caught between early promise and meaningful impact, the instinct is often to search for better tools, more data, or more advanced models. In practice, the blockage usually sits elsewhere.


It appears where AI meets real decision making.


Trading decisions are time sensitive, commercially delicate, and shaped by context that rarely fits neatly inside systems. Introducing AI into this environment raises questions that extend beyond performance metrics. How much trust is sufficient. Who carries accountability when outputs are incomplete or wrong. When human judgement should override automated insight, and how that intervention is recorded.


Many organisations realise at this stage that they have never fully described how decisions are made today. Processes exist, but their interpretation varies across desks and regions. Exceptions are handled skilfully, but informally, and risk is managed through experience rather than explicit frameworks.


AI struggles with ambiguity of this kind, and while it does not fail spectacularly, it simply cannot advance without clearer boundaries. This is where many initiatives slow dramatically, not for lack of potential, but because they have reached the limits of implicit understanding.


BRITTLENESS IS THE HIDDEN RISK


As AI programmes expand, another risk tends to develop. Infrastructure that looks advanced on the surface can become fragile underneath.


Brittleness rarely stems from a single poor choice, tending to grow through a series of sensible short term decisions. Whether that’s a POC being pushed into production with minimal redesign, or a model becoming tightly coupled to a specific data feed. Validation is layered on afterwards. Governance is attached rather than engineered.


Gradually, they end up with systems that function well only while conditions remain stable and integrating new data sources becomes costly. Changing tools feels hazardous, and improving models threatens downstream processes that were never designed to rely on them.


In a field where markets and technology both evolve quickly, this creates a difficult paradox. AI is introduced to increase agility, yet the organisation becomes less able to adapt.


AI DOES NOT REWARD SPEED FOR ITS OWN SAKE. IT REWARDS HONESTY ABOUT READINESS.


Most importantly, they define AI’s role in decision making before scaling its reach. They are clear about where it informs, where it recommends, and where it acts and human accountability is explicit, not assumed.


REDEFINING PROGRESS


Success with AI in energy and commodities trading is often described in terms of sophistication. Smarter models. Faster computation. Greater automation.


In practice, success is more subtle than this.


It appears in organisations that can defend their numbers under pressure, and in systems that evolve without undermining trust. Teams that succeed are the ones that understand not only what the technology is doing, but why.


AI does not reward speed for its own sake, nor does it penalise caution. It rewards honesty about readiness.


AI is not replacing traders. It is exposing how trading organisations truly function, and those prepared to face that reality will find progress comes more naturally. Those who are not will continue to mistake activity for momentum.


LAURENCE PISANI Energies and Commodities Practice Director laurencep@digiterre.com


6 | ADMISI - The Ghost In The Machine | Q1 Edition 2026


Page 1  |  Page 2  |  Page 3  |  Page 4  |  Page 5  |  Page 6  |  Page 7  |  Page 8  |  Page 9  |  Page 10  |  Page 11  |  Page 12  |  Page 13  |  Page 14  |  Page 15  |  Page 16  |  Page 17  |  Page 18  |  Page 19  |  Page 20  |  Page 21  |  Page 22  |  Page 23  |  Page 24  |  Page 25  |  Page 26  |  Page 27  |  Page 28  |  Page 29  |  Page 30  |  Page 31  |  Page 32  |  Page 33  |  Page 34  |  Page 35  |  Page 36  |  Page 37  |  Page 38  |  Page 39  |  Page 40