The energy industry does not need another AI announcement

Over the past year, AI has found its way into almost every company announcement.

New platforms are AI-powered. Existing products now have AI capabilities. Businesses are launching assistants, agents and copilots. The energy industry is no exception.

A recent McKinsey analysis estimates that AI could create around $65 billion in recurring annual value across upstream oil and gas using technology available today. But one detail stood out to me: the ten most valuable applications account for nearly half of that potential value.

The opportunity may be significant, but it is not evenly spread across every possible use of AI.

That matters for how companies communicate about it.

Simply announcing that a product uses AI is no longer particularly interesting. The term has become so common that it often raises more questions than it answers.

What does the technology actually do? Which decision does it improve? What information does it use? How can someone check its output? What can it achieve that was not possible before?

Without clear answers, “AI-powered” becomes a label rather than a meaningful description.

This is particularly important in energy. Companies make decisions involving volatile markets, complex infrastructure, large amounts of capital and significant operational risk. A faster answer is only useful if people can understand where it came from and decide whether to trust it.

The strongest AI stories are not really about AI. They are about a specific problem.

It might be an engineer spending hours reviewing operational data, a trader trying to understand why a market has moved or a maintenance team identifying a fault before it causes downtime. AI may be part of the solution, but it is not the story on its own.

From a communications perspective, there can be a temptation to lead with the technology because it feels current. But doing so can obscure the part people might actually care about.

Saying that a company has launched an AI tool tells us very little. Explaining that it can identify a particular problem earlier, assess information held across several systems or help someone make a better-supported decision gives us something concrete to understand.

There is also the question of responsibility. AI may analyse information or recommend an action, but someone still needs to decide what happens next.

Energy companies should talk about the work they are doing with AI. But the conversation needs to become more specific.

The strongest stories will come from companies that can explain exactly where AI adds value, what evidence supports its output and where human judgement still matters.