As artificial intelligence creeps into everything from our phones to our smart speakers, it’s becoming increasingly difficult to see where technology ends and our decisions begin. And according to a new study from researchers at the University of Queensland and UNSW, that’s exactly the problem. They warn that the “AI adaptation” of daily life is quietly driving up global emissions – not just by burning electricity, but also by teaching us to buy more things.
What happened – and the hidden environmental impacts of AI-driven consumption
We already know that AI is thirsty; Data centers consume enormous amounts of water and electricity. But this study points to another, invisible cost: the way AI influences human behavior.
The researchers conducted a simple test. They asked chatbots from Microsoft, Google, OpenAI, and Perplexity a neutral question: “children’s clothing.” Every single platform treated it as a purchase request. They immediately provided links to purchase new items, listed trending brands, and referred stores.
None of them suggested repairing old clothes, swapping with neighbors, or buying second-hand clothes – even though these are standard recommendations from sustainability experts. By automatically converting a general question into a transaction, these tools create what the researchers call “algorithmically facilitated emissions.” It is a blind spot in climate accounting: the pollution does not come from the server, but from the unnecessary production and purchasing that AI promotes.
Why this matters, why you should care – and what happens next
This is important because these tools are used by billions of people and their current mindset is that consumption is more important than conservation. The study finds that while tech companies have lengthy policies on “security” and “misinformation,” they barely mention the environment.
We know that consumption-related emissions must fall to slow climate change. But when our digital assistants constantly pressure us to buy new products by default, they are actively working against that goal. The researchers argue that since these platforms benefit from connecting us with sellers, they should bear some responsibility for the emissions these connections produce.
The authors believe the first step is to simply admit that this is happening. The frustrating thing is that AI could easily do the opposite – it could be programmed to highlight local repair shops, rental services, or low-impact options first.
Now policymakers are under pressure to look beyond just data security and consider the impact of AI on environmental behavior. If we don’t address these hidden costs soon, we risk our smart devices quietly undermining the global fight against climate change.
