AI hasn’t suddenly arrived in organisations, but its role has shifted quickly.
What was once applied in targeted ways is now becoming central to how products are built, and how work gets done.
At Culture Amp, we’ve been using machine learning for more than a decade. Features such as sentiment analysis and topic clustering in employee engagement surveys relied on models we built ourselves and applied in contained ways where they added clear value.
What’s changed over the past year is how central AI has become.
According to Deloitte’s 2026 State of AI in the Enterprise, workforce access to AI has expanded by 50% in just one year.
It’s moved from being used in specific parts of a product to sitting at the core of how products are built and experienced. Capabilities like AI Coach now flexibly support managers by meeting them where they are, rather than managers having to navigate rigid structured workflows. Beyond product development, AI is being used across almost every role as a productivity tool.
Get the best of Startup Daily straight to your inbox
Want to know the latest in startup news? Subscribe to our daily news and analysis coverage on what’s happening to ANZ startups, investors and the broader ecosystem. And best of all, it’s FREE!
By continuing, you agree to our Terms & Conditions and Privacy Policy.
It’s not just adoption that’s changed, it’s also the scale it’s reached. And at this point, AI stops behaving like a feature and starts behaving more like a flexible foundation, with value creation that extends beyond individual use cases.
This shift is happening against a broader backdrop. The global energy transition is still incomplete, and the rapid growth in AI is increasing demand on data centres that require significant energy. That makes how AI is used, and how quickly it scales, a more consequential set of decisions for business leaders.
The next challenge in AI isn’t capability, it’s responsibility
The instinct is still to treat AI as software: something you introduce, test and layer into workflows. But the more it’s used, the more it behaves like a system running underneath the business.
It draws on shared, energy-intensive infrastructure and scales with usage in ways that aren’t always obvious, and this includes carbon.
The carbon impact of AI scales with how widely it’s used across the business, and how efficiently or otherwise it is deployed.
This shift is happening against a broader backdrop. The global energy transition to renewables remains incomplete, and rising demand for AI is increasing pressure on energy-intensive data infrastructure.
The International Energy Agency’s latest analysis projects that electricity demand from data centres, AI and digital infrastructure will grow rapidly over the next few years, placing additional strain on energy systems already under pressure.
That makes how AI is used, and how quickly it scales, a mor consequential set of decisions.
Measurement is lagging behind adoption
For most organisations, this isn’t entirely new. The majority of emissions in technology businesses already sit in cloud infrastructure and data centres, outside direct control and often with limited transparency. AI is increasing demand on those systems without making that impact easier to see.
At Culture Amp, this is something we’ve been working through as part of our broader sustainability efforts. As a certified B Corp, we approach this through a broader lens of responsibility, balancing innovation with accountability to our employees, customers and the wider environment.
One of the more encouraging things we’ve seen is that improving efficiency, reducing cost and lowering emissions often point in the same direction.
By making targeted changes to our cloud architecture and usage patterns, we reduced downstream data centre emissions by 49 per cent while also lowering operating costs.
That experience doesn’t map directly to AI, but it does shape how we think about it.
In most companies measurement remains underdeveloped, but it is a prerequisite for responsible deployment. You cannot optimise what you cannot see.
Where your systems run matters more than you think
There are still gaps. There’s no standard way to attribute emissions to different AI use cases. Visibility into vendor infrastructure is improving, but still limited. And most organisations are only starting to understand how usage patterns: from model selection to real-time versus asynchronous workloads, translate into impact at scale.
What it does point to is the need to treat this as a system that needs to be actively managed.
At scale, these decisions don’t just sit within individual organisations. They shape demand across shared infrastructure and, over time, the systems that support it.
What this requires from organisations
Organisations don’t need perfect data to start making better decisions but they do need to be deliberate about how AI is used and scaled.
- Treat carbon like a cost to be managed
Think about carbon in the same way you would a P&L. At Culture Amp, we’ve worked to understand our footprint across the life of the company and treat it as something to be actively managed over time — minimising adding to our debt by operational improvements, and then paying down our debt through the purchase of carbon removal credits (not just carbon offset credits).
For most businesses, the first step is simply to start measuring and taking accountability for your impact.
- Build visibility into AI usage and cost
AI usage has real cost signals. There’s a close relationship between how AI is used, the number of tokens consumed and the cost of running those systems.
The cost is not just in dollars but also in carbon, and without a global price on carbon it is beholden on responsible companies to ensure the carbon impact is priced in.
Creating visibility into where AI is being used, and holding teams accountable for usage and infrastructure costs, is one of the most practical ways to manage both.
- Match the model to the problem
Not every task requires the most advanced model, and some might not even need AI at all. More complex reasoning models are more resource-intensive, and in many cases smaller models will deliver the same outcome.
Being deliberate about model choice can materially reduce cost with no loss of business value.
- Design for efficiency at the architecture level
How systems are built matters.
Engineering decisions, from how workloads are structured to how systems scale, have a direct impact on both cost and emissions. Investing in efficient architecture early compounds over time.
- Choose partners with clear commitments
For most organisations, infrastructure is external. That makes vendor choice important.
Cloud providers are increasingly making commitments around water positivity, renewable energy and net zero targets, and those commitments should form part of how organisations think about where their systems run.
Responsible AI is part of building better companies
As AI becomes embedded across industries, the question isn’t just how it’s used but how it’s governed, measured and sustained over time.
AI is no longer something organisations are experimenting with. It’s becoming part of how work gets done — and, increasingly, part of the systems that underpin it.
It’s no longer just about what these systems can do, but how they’re run, measured and sustained over time. And in many organisations, adoption is moving faster than the ability to manage what’s being built.
At scale, that doesn’t just affect individual businesses. It shapes demand across shared infrastructure and the systems that support it.
At Culture Amp, we’ve always taken a people-centred approach to how work gets done. AI should strengthen that by helping people make more informed, more timely and better decisions and hence be more effective in their roles, not replacing them.
The organisations that get this right will be the ones that build visibility early, make deliberate decisions about how and where AI is used, and treat it as something that needs to be actively managed over time.
That’s what will ultimately define not just performance and cost, but whether these systems make work better for people and are sustainable at the scale they’re now being built.
- Doug English is Chief Technology Officer & Co-Founder at Culture Amp.
