TRUST WILL DEFINE SUCCESS
Deploying agentic AI within a commodities organisation presents challenges that extend well beyond the technology itself. These businesses manage commercially sensitive information, operate within heavily regulated markets and depend upon rigorous governance.
An agent must understand permissions. It must provide an auditable account of the information, assumptions and actions that produced its recommendation. It must reference its sources. It must recognise uncertainty and know when a human decision is required.
Perhaps most importantly, organisations must resist the temptation to pursue AI for its own sake.
Successful adoption will depend less on selecting the latest foundation model and more on building a trusted operating environment around it, combining enterprise data, robust architecture, security, governance, testing and human oversight.
Experimentation can demonstrate what is possible. Moving agentic systems into production requires proof that they are reliable, secure and genuinely valuable within real business workflows.
For the first time, technology is beginning to bridge the gap between information and action and companies that embrace this thoughtfully will not necessarily employ fewer experts. Instead, they will enable those experts to make faster, better-informed decisions with significantly less effort.
In an industry where competitive advantage has always depended upon the quality and speed of decision-making, this may prove to be one of the most important technological advances yet.
FOR THE FIRST TIME, TECHNOLOGY IS BEGINNING TO BRIDGE THE GAP BETWEEN INFORMATION AND ACTION.
A QUIET REVOLUTION
The history of commodities trading has been characterised by incremental technological change rather than overnight disruption. Electronic trading supplemented open outcry. Algorithmic pricing enhanced decision-making. Cloud platforms replaced ageing infrastructure. Advanced analytics improved forecasting.
Agentic AI feels different.
Not because it replaces people, but because it changes how people interact with the vast digital ecosystems they have spent decades building.
For decades, invaluable human expertise has been scattered across systems, documents, inboxes and spreadsheets.
Agentic AI offers the possibility of bringing that knowledge together, not to replace human judgement, but to put it to work.
Laurence Pisani Energy & Commodities Practice Director E.
info@digiterre.com
15 | ADMISI - The Ghost In The Machine | Q3 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
