Shape the AI revolution: understanding our new AI Ethics and Society MSc
Artificial Intelligence (AI) is being tested and adopted at pace across practically every industry, reshaping the world around us. But behind the hype and disruption of this technological revolution, there are still many serious questions surrounding the ethics of AI and how we, as a society, want it to function.
The new MSc in AI, Ethics and Society, run by the Leeds Institute for Societal Futures, aims to train experts to navigate the topic and provide solutions that ensure AI has a positive and responsible impact.
To shine a light on the course, why it’s relevant and what you could gain from studying it, we sat down with Programme Lead Dr Andrew Kirton, and Module Leads Dr Jenny Sexton, Dr Dibyadyuti Roy, and Dr Linhao Fang.
Why do ethics and society matter to AI today?
When we consider what makes AI, our thoughts often jump to the technologies that underpin it – things like machine learning, large language models, deep learning, and generative AI. However, this ignores the fact that all these systems are the product of human thinking and decisions.
“AI is only human”, explains Dibyadyuti, “and if I want to point out why ethics is important, it’s because as humans, what is the core of the word or the etymology of the word ethics? It goes back to morals. Morality is a crucial, universally valued human concept. So, therefore, you can’t do AI without ethics.”
The ethics of AI also have profound consequences for the role of labour in society
Linhao agrees: “It is tempting to fall into the false dichotomy between technology versus human. The impulse to develop technology has always been a defining feature of our species, whether that’s a stone tool or a large language model. Like Latour puts it, it is “society made durable”. Many of the patterns we see in emerging technologies such as AI may appear novel, but the mechanisms behind them are not.”
Many of today’s AI applications are being shaped by humans, often people in less economically developed countries, so-called ‘invisible labour’. If we want our AI applications to have a positive ethical footprint, these issues must be addressed as well.
“Data annotation for autonomous driving systems happens in my hometown in Calcutta. Typically, this labour forms about 80% of any algorithmic model work that shapes modern AI systems”, explains Dibyadyuti. “A major tech retailer launched a chain of cashier-less stores, promising customers a fully automated shopping experience. But behind the scenes, there was no AI at all — instead, roughly a thousand workers in India were monitoring security camera feeds and manually assembling virtual shopping carts
"The ethics of AI also have profound consequences for the role of labour in society,” according to Andrew. “The push for Large Language Models (LLMs) to take on knowledge‑work roles is transforming white‑collar jobs in the same way gig work transformed taxi driving.”
“If experts like scientists and lawyers are losing their jobs, and the only work they can find is training AI models to do that work, it transforms how society understands and treats expertise. If, in the future, experts are not someone you have contact with, and it’s all just people updating LLMs, that’s a massive change to our societal setup.”
Approaching AI from a purely technical standpoint also has significant ethical implications, as Jenny explains:
“We’re already seeing lots of small‑scale failures. We need to be thinking about things like hiring algorithms that chuck out everyone over a certain age, or medical applications that don’t work for groups that weren't in the initial study. All these things are technical failures with ethical consequences. But these things are predictable if you understand the limits of statistics and data quality and the algorithms underneath.”
How is this course relevant to today’s AI landscape?
We’ve discussed a snapshot of the issues the new course explores. But as AI advances and improves at a lightning pace, can prospective students trust that their education will remain relevant after they graduate?
For Andrew, the course is enduring due to the challenges it focuses on. “Each module tackles a fundamental or foundational problem – the kind of problem that will be long‑standing based on the way this technology works. Bias, injustice, accountability, global inequalities… these are chunky problems that will never go away, even as the technology develops.”
The course is one of the first developed at Leeds that spans multiple subject areas. As Linhao explains, “Interdisciplinarity is key. People are often highly specialised in their own fields, and it’s easy to work in silos. We tend to focus on the few KPIs that matter to us instead of considering society as a whole. Ethics deals with normative questions, and such questions rarely have definitive right or wrong answers. They require collective deliberation across different groups and all walks of life.”
The course was co-designed with input from students, academics, and industry in ‘sandpit sessions’. “I had really surprising conversations”, explains Jenny. “They tuned me into how we might need to teach this. Students and industry partners came up with ideas outside the structure I’d considered. It helped me identify how to build the foundations students need.”
Dibyadyuti agrees: “The sandpit did not allow for any context rot [the phenomenon where LLMs provide less accurate answers as context provided grows too large]. The context kept getting updated and adapted through every new conversation. It forced us to think about the programme in a way that was dynamic and responsive.”
How does the course prepare students for their future careers?
The new course has been specifically designed for students aiming for roles in AI leadership, consultancy, research, and analysis. What skills will students learn to support them in these careers?
Understanding and navigating real-world complexity is an overarching focus. “What you have in the real world are broad problems – not problems unique to particular disciplines”, says Andrew. “The course prepares students from different disciplinary backgrounds to deal with that.”
The strength of this programme is the fact that it brings in scholars, teachers, pedagogues, researchers, and industry” says Dibyadyuti. “This allows you to appreciate that AI is not one thing. At the core of it is human innovation, labour and accountability.
Bridging technical and ethical thinking is also a key part of the course. “AI engineers aren’t usually taught this. Education can place too much focus on the technical side and too little on interpreting the effects of the work. We want students to understand the limits of statistics, data quality, algorithms – and how those limits translate into ethical consequences”, explains Jenny.
The course’s interdisciplinary approach also applies to global cultural and political differences related to AI, giving students skills valued highly by global industry.
As Linhao notes, “Attitudes towards AI are not shared evenly across different cultural and political contexts. The development of contemporary large foundation models demands infrastructure and capital on a scale so vast that it often becomes a national-level endeavour. This makes the issues involved inevitably geopolitical, and public narratives around AI can be heavily shaped by governmental strategic priorities. People from different backgrounds may hold very different assumptions, expectations, and preferences regarding AI, and we want students to be able to recognise and navigate this complexity.”
Taking studies a step further, you can also take part in the University’s Global Industry programme's professional opportunities to apply your learning in a real-world context.
Like all new technologies, the introduction of AI requires serious consideration and guidance. If you want to influence the direction of AI and ensure it works fairly for society, take a look at our new course. If you have any questions, get in touch with our team.