In the autumn of 2022, several of Europe’s largest energy companies came close to failure — not because their trading strategies were wrong, but because they were right. As gas and power prices spiked, utilities that had prudently sold their output forward faced enormous variation margin calls on those hedges: collateral demands so large that some needed emergency liquidity backstops from governments and banks to stay solvent. Two years earlier, WTI crude futures had done something no model had priced as plausible — they traded below zero, leaving holders of long positions paying to dispose of oil they could not store.
Neither episode was an information problem. The data existed. What failed was the ability to see exposure building across a portfolio, to size it against liquidity and credit limits, and to act before the market moved against the position.
That gap — between knowing and deciding under pressure — is where large language models (LLMs) are now reshaping the economics of risk management (RM) consulting.
Traditional risk advisory work — document review, regulatory synthesis, compliance mapping, and report generation — can increasingly be automated or dramatically accelerated by AI. For a consulting model built on the billable hours of gathering and organizing information, that is a structural change, not a marginal one.
But for energy companies active in trading and contracting, access to information was never the binding constraint. The hard problems live in the exposures themselves: physical supply contracts, counterparty credit, and large derivative positions in futures, forwards, swaps, and options whose combined risk shifts by the hour. Managing that book — where a single mispriced correlation or an unnoticed concentration of counterparty exposure can be measured in billions — is a matter of judgment, not retrieval.
The value is migrating
This is why the value of consulting is shifting away from information gathering and toward decision enablement.
Historically, risk consulting relied on large teams performing periodic assessments and producing static reports — a snapshot of exposure often out of date before it was delivered. For a trading organization, a quarterly view of risk is close to useless; positions, prices, and creditworthiness change continuously. Energy clients will increasingly expect continuous intelligence, integrated exposure monitoring, and AI-enabled decision support that keeps pace with the market.
The firms best positioned for this transition will act less as traditional advisors and more as “risk systems integrators,” helping clients combine AI capabilities with trading judgment, governance, and enterprise-wide risk visibility. AI alone cannot set a credit limit, price a complex structured contract, or decide how much basis risk a book should carry. Those decisions still require human judgment and deep market expertise. The consultant’s role is to ensure the technology sharpens that judgment rather than obscuring it.
Three priorities for the transition
First, build AI-augmented exposure sensing. Firms should help clients continuously monitor market and credit signals across the trading book — mark-to-market moves, margin and collateral requirements, concentration and position limits, and the early deterioration of a counterparty’s credit. The value lies in catching the anomaly while it is still small: a counterparty whose financials are weakening well before it defaults, a basis relationship drifting before it breaks, a position quietly breaching its limit before it becomes a headline loss.
Second, deepen portfolio stress testing. Value-at-Risk and single-factor sensitivities describe the world on ordinary days; they tend to fail precisely when it matters. Energy risks are correlated and non-linear — price, volatility, and basis move together; correlations that diversified a book in calm markets collapse in a crisis; and “wrong-way risk” emerges when a counterparty’s creditworthiness is itself tied to the very price move that puts the trade in the money. LLM-augmented tools make it feasible to construct and explore a far wider range of combined scenarios: a price spike that simultaneously triggers margin calls, drains liquidity, and impairs the counterparties on the other side of the hedge. That is how risk actually arrives — rarely one factor at a time.
Third, govern the models, not just the trades. Trading organizations already live with model risk; AI raises the stakes. Before a firm relies on an AI-enabled system to value contracts, flag exposures, or inform hedging decisions in high-consequence environments, it needs transparency, auditability, cybersecurity controls, and meaningful human oversight — the same discipline of model validation and control that regulators already expect of any pricing or risk engine. Helping clients answer “can we trust this model, and can we prove it to a regulator or an auditor?” may become one of the most valuable services a firm offers, and one where energy-market expertise is hardest to replicate.
A new basis for competition
Consulting firms will increasingly compete on their ability to combine four things: energy-market and trading expertise, a working understanding of physical and financial operations, AI fluency, and systems thinking. Any one of these in isolation is becoming a commodity. Their intersection — someone who understands both a swap’s collateral mechanics and a model’s failure modes — is not.
The traditional model, built around manual analysis and static deliverables, is losing value precisely because the manual work is the part machines now do well. What remains scarce is the capacity to design systems that produce continuous intelligence, strengthen organizational resilience, and support better decisions under uncertainty.
None of this diminishes the human element; it makes it more central. When routine analysis is automated, the differentiator is judgment — knowing which exposures matter, and quantifying the cost of these risks to the organization. The best firms will use AI to elevate that judgment, not to substitute for it.
LLMs will not eliminate risk management consulting in the energy sector. But they will redefine where firms create value — moving it decisively from the report to the decision, and from the periodic assessment to the living, position-aware system. The advisors who understand that shift, and rebuild their offerings around it, will define the next era of the profession.
