Don’t rely on AI

Have you used Generative AI (GenAI) to design chemistry practicals? For researchers in education, this is still new territory, but some teachers are already experimenting. Matthew Gundry, RSC excellence in secondary and further education 2025 prize winner, shares: ‘I don’t rely on it fully, but sometimes it throws something back at me I hadn’t considered.’

And for Aidan Simcock, a recently qualified science teacher, GenAI is a game-changer for accessibility. He uses it to translate practical instructions: ‘Pair the English text with the instructions in the student’s first language and add stepwise images,’ he explains. ‘It perfectly dual codes the instructions for EAL learners while still supporting the rest of the class.’

Without human intervention, AI can get it wrong. But to stay competitive in modern chemistry research, using it is unavoidable

Elena Mecklenburgh, a chemical engineering undergraduate at Leeds University, reveals, ‘I often find myself using GenAI to help plan out the structure of my lab reports, which can be very helpful, but I also find it can limit your creative thinking.’ Tim adds, ‘GenAI is encouraged as a tool for collaboration and critical thinking, but we strongly discourage its use as a shortcut that bypasses learning.’

The challenge is ensuring everyone understands the chemistry and the risk rather than simply copying AI output. Without human intervention, AI can get it wrong. But to stay competitive in modern chemistry research, using it is unavoidable. Dr David Bray, team leader for chemistry and materials at STFC Hartree Centre, emphasises to me that a public understanding of how science and technology drive environmental impact is crucial, empowering people to become ‘critical consumers’ of AI. He explains that a ‘digital first’ approach supported by AI is much more sustainable than traditional trial-and-error methods.

AI is reshaping how industrial laboratories and research operates. Yet this computational power cuts both ways, raising important questions about practical skills and future careers. Will self-driving laboratories represent the future? Bob asks: ‘Does AI have the nuance of fifty-plus years of hands-on experience, and can it invent new techniques?’ The evidence suggests clear limitations. Matthew tells me, ‘I once got GenAI to help me write a method to plan an iodine clock reaction so solutions would change colour at specific times, but the practical failed miserably.’

Will AI-driven, robot-led laboratories come with a cost? Profit could take precedence over people. If we allow technology to advance without safeguards, we risk creating a future where innovation thrives while human expertise is cast aside. Ethical responsibility must keep pace with automation.

As AI transforms how chemistry is conducted in research and industry, the skills learners will need for their future careers are changing, too. They will need to critically evaluate, interrogate and collaborate with AI tools. So, as we embrace AI in chemistry education, policy makers and educators must maintain rigorous safety standards, preserve the irreplaceable value of practical skills, teach students to be critical consumers of AI-generated information, ensure equity of access and celebrate human creativity alongside what AI makes possible. We cannot afford to let AI replace the unpredictable nature of chemistry; instead, we should use it to sharpen our students’ curiosity, critical thinking and analytical skills.