The AI Revolution: Where Innovation Clashes with Ethics.

Written By Karen Daniels

AI is no longer a future conversation, it’s a present-day business advantage.

From automating operations to accelerating decision-making, innovation is moving faster than most organisations can absorb. For SMEs especially, the promise is clear: do more, faster, and at lower cost.

But beneath that momentum lies a growing tension.

Because as businesses race to adopt AI, few are stopping to ask the harder questions:
What should we be doing? not just what can we do?

This is where innovation begins to clash with ethics.

And for organisations that pride themselves on being responsible, sustainable, and forward-thinking, that clash isn’t something you can ignore, it’s something you need to navigate strategically.

The technology itself isn’t a problem, it’s the size, locations and speed at which the data centres being built. In the last 5-ish years there’s been growing momentum on environmental and social issues within the business world and now we have a rapidly growing conflict to that effort .

AI is creating a dilemma for ethical and responsible businesses.

It is a resource intensive technology so it’s no wonder that those who are looking to build them are seeking locations with poor environmental regulation, but these regions are the same ones with:

  • water scarcity

  • poor grid connectivity

  • Poorer air quality

  • lower quality of life

…and yet they’re the most affected by climate change.

AI data centres in these regions will lead to worsening the conditions for the people who live there so what can we do as an ethical or responsible business? How do we keep pace with the changing world without hurting it?

We’re at the beginning of AI governance; the technology has moved so rapidly that legislation hasn’t had chance to catch up yet. In the UK currently there is zero AI specific regulation and is, instead, using a ‘Pro-innovation, principles-based framework’, the 5 core principles are:

  • Safety, security and robustness

  • Transparency and explainability

  • Fairness

  • Accountability and governance

  • Contestability and redress

This all sounds nice, but enforcement is unlikely, existing (already overstretched) regulators are expected to utilise existing laws, not drafted with AI in mind, to regulate AI within their industries. So that leaves us relying on business to self-regulate and implement ethical AI use and governance practices. Which sounds like a big ask, but there is starting to be some structure emerging.

ISO 42001 is one of the first real attempts to formalise what responsible or ethical AI management looks like at an organisational level. Climate change is now technically built into all ISO management system standards, including ISO 42001. But the requirement is only to consider whether it’s relevant, not to actively manage or reduce impact. So even with a formal standard in place, it still comes down to whether a business chooses to treat it as important.

It’s important to recognise that this is not about the technology itself, it’s about the management system around it. So it’s similar to ISO 14001 for environmental management or ISO 9001 for quality, ISO 42001 focuses on putting governance, accountability and oversight in place for how AI is used.

At its core, it’s asking businesses to do a few important things:

  • be clear on where and how AI is being used in the organisation

  • assess risks, not just operational but ethical and societal

  • put controls in place to manage those risks

  • assign responsibility and oversight, rather than leaving it vague

  • and continuously review and improve as the technology and its impacts evolve

It doesn’t solve the infrastructure problem. It doesn’t fix the environmental impact of data centres.

But it does do something important. It moves AI out of the “just implement it” space and into the same kind of structured, accountable thinking we apply to other critical areas of the business.

And that matters.

Because if regulation is still catching up, then frameworks like this are one of the few ways businesses can demonstrate that they’re taking this seriously. Not just from a compliance point of view, but from a credibility one.

For businesses that already care about sustainability and responsibility, this isn’t a stretch. It’s an extension. You’re already thinking about supply chains, environmental impact and long term risk. You’re already asking questions about where things come from, who they impact and what the consequences are beyond your immediate business.

That’s the bit we need to carry across into AI because you can’t make responsible decisions about AI if you don’t understand what’s sitting behind it. Where it’s hosted. What it’s consuming. What impact it’s having. Is it water or air cooled - keep asking the questions!

Having just enough visibility to understand the impact, not only the outcome. AI in itself doesn’t change that, if anything, it makes it more important because when the impacts are less visible, it becomes easier not to think about them.

Frameworks like ISO 42001 start to give some structure to this, but that doesn’t mean SMEs need to go through certification to do this well. The principles are accessible. Understanding where you’re using AI, thinking about the risks, being clear on responsibility and making conscious decisions about how and when you use it.

That can all be done without significant cost or complexity.

This is just applying that same thinking to a different part of the business and maybe that’s the shift that needs to happen.

Not treating AI as something separate or exceptional, but treating it like any other business system that has risk, impact and consequences, because the question hasn’t changed.

It’s still not just about what can we do.

It’s whether we’ve thought through what we should do, and whether we’re willing to stand behind that.

Karen Daniels