What is AI, how does it work and why are some people concerned about it?
BBC News ● Covered by 45 sources
AI is everywhere now, from feeds to chatbots to medicine. People worry it can spread bias, falsehoods and use too much power and water.
Based on reporting by BBC News — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
AI has quietly moved from buzzword to plumbing. It helps sort social feeds, suggest music, power voice assistants and even support medical work like spotting cancers, reading X-rays and speeding diagnoses. The same tech is also behind the chatbots that have shoved AI into the centre of public argument.
At its simplest, AI lets computers chew through huge piles of data, spot patterns and act on detailed instructions. It does not think or empathise. Instead, scientists build systems that imitate parts of how people learn and use knowledge. That can mean recommending a product based on past purchases, or helping self-driving cars make sense of the road.
The loudest attention now sits on generative AI. These tools learn from vast amounts of existing text and images, then produce new material that can look human-made. ChatGPT, DeepSeek’s chatbot, Google’s Gemini and Meta AI can handle text; Midjourney and Veo 3 are built for images and video. AI can even make music, which is how songs that mimic famous artists end up confusing listeners about what’s real.
But the backlash is not just about who gets replaced. Experts keep pointing to current failures: biased outputs, false claims presented as fact, and even made-up sources. Apple pulled one AI feature in early 2025 after it mangled news notifications, and Google drew criticism for errors in its AI search summaries. Schools, workplaces and creators have all pushed back too, with artists accusing developers of training on their work without consent or pay.
Then there’s the physical cost. No one really knows the industry’s full energy use, but some researchers think it could soon match the Netherlands. AI also needs more data centres, more chips, more power and often more water. Governments are responding in very different ways: the EU has a formal law, China ties rules to censorship and data controls, while the UK has gone for narrower bans and safety testing instead of a big new AI statute.
My take — AI-written commentary, not fact-checked reporting
The real story is not that AI is magical; it’s that the costs are turning out to be gloriously unmagical. The industry keeps selling future shock while asking everyone else to eat the power bill, the water bill and the legal risk. That’s not innovation, that’s a tab with better branding.
Read more about this at: BBC News
