The new age of insight? Questioning bias in an AI-powered world

I probably shouldn't admit this as a researcher who spends their days dealing in facts and evidence, but when it comes to reading for pleasure, I’ll often opt for the fantasy world of a novel over a hard-hitting work of non-fiction. However, when I spotted Laura Bates’ ‘The New Age of Sexism: How the AI Revolution is Reinventing Misogyny’ in my local bookshop, I felt compelled to make an exception. 

Much of the discourse about AI today understandably focuses on its benefits. It promises to save time, boost productivity, and help us process information at a previously unimaginable scale. Yet Bates’ book asks us to confront an important question that receives far less attention: how are these systems being built, and whose perspectives are reflected within them? 

AI systems are trained on enormous volumes of human-generated content, inevitably reflecting the structural dynamics and inequalities already present in our society. Biases relating to characteristics such as gender, ethnicity, age, and class can therefore be reproduced in AI outputs, often in ways that are difficult to detect. The pace of AI development has also, at times, outstripped the regulatory and ethical frameworks designed to govern it.

This has significant implications for conducting research. As an industry, we are increasingly embracing AI across the project lifecycle, using it to brainstorm hypotheses, craft research instruments, analyse large volumes of data, and develop key findings. The efficiencies are hard to ignore. However, the more embedded these tools become in our workflows, the more important it is to understand the assumptions and biases that lurk beneath the surface.

This matters because good research is not simply about finding answers. It is about uncovering complexity, challenging assumptions, and identifying perspectives that may otherwise be overlooked. If we rely too heavily on systems trained on existing patterns of thought, we risk perpetuating the very same biases that we seek to challenge.

As Bates puts it: ‘We must recognise that if [AI] technologies are the great project of this age of humanity, our aspirations for them must be exactingly high…We can continue to mine the opportunities presented by AI to make life better for billions of people, while also building its framework to safeguard against bias, unequal outcomes, and the replication of existing prejudices’ [1].

For researchers, this doesn't mean rejecting AI entirely. It means using it more consciously, interrogating its outputs, and applying context and human judgement at every stage of the process. AI may help us work more efficiently, but it is our responsibility to ensure that the insights we deliver are robust, meaningful, and genuinely reflective of the people we strive to understand.

[1] Bates, L. 2025. The New Age of Sexism: How the AI Revolution is Reinventing Misogyny. London: Simon & Schuster. (p.268).