ARTIFICIAL INTELLIGENCE
AI can no longer be compartmentalised if graduates are to be adequately prepared for the forces reshaping business and society. Imperial Business School’s Tommaso Valletti outlines the process of co-ownership and collaboration behind a new master’s degree that combines technical AI training with economics, law and governance
That was the starting premise for the MSc in AI, Economics & Policy at Imperial Business School, which has just opened for applications ahead of an inaugural cohort start date of September 2027. Delivered jointly by Imperial Business School and Imperial’s Department of Computing, it is the UK’s first master’s degree to combine technical AI training with economics, law and governance in a single curriculum. Identifying the root of the problem I’ve spent much of my career working on competition economics and digital markets. That includes time spent engaging directly with regulators grappling with how to apply decades-old frameworks to new markets when, for three years, I took leave from academia to serve as chief competition economist for the European Union. What became clear, over and over, is that the people best placed to shape good tech policy are rarely the people who best understand how the underlying systems work – and vice versa. The same problem exists inside companies: the people making decisions about adopting or governing AI may not understand the technology well enough, while those developing it may not fully understand the wider business, economic and societal consequences of what they build. Regulators need people who can read a model card and a market study with equal fluency. Companies need people who can make decisions about AI and understand both the economics of digital markets and the technical realities of the AI systems they’re assessing. Consultancies, meanwhile, need people who can move between a boardroom conversation about competitive strategy and a technical conversation about model capabilities without the use of a translator. That’s a very specific skills gap and it’s one that neither a pure computing degree nor a pure economics or policy degree closes on its own. So, the only honest way to build the new MSc programme was to build it jointly, rather than to have Imperial Business School design most of it and bolt on a few computing modules, or vice versa. We therefore pursued a path of real co-ownership all the way from curriculum design through to teaching delivery. Integration beyond marketing lines “Cross-disciplinary” is a word that now appears on almost every prospectus and it usually means students take a handful of electives outside their home department. We wanted something different – a curriculum where the AI systems teaching and the
A rtificial intelligence is Moreover, companies are making decisions every day about how to use AI, whether it will genuinely create value, how it will change the competitive landscape and what risks it may create. Boards are making multi-year strategic bets on a technology that changes shape every few months. Competition authorities are being asked to police market power in industries that didn’t exist five years ago. Financial regulators are assessing systemic risk in models few of their staff fully understand. Governments everywhere are trying to write policy for a moving target. Universities sit inside that same pressure, whether we like it or not. For a long time, higher education could afford to treat AI as a technical subject taught in computer science departments, with economics, law and public policy departments occasionally rewriting how companies compete, how work gets done and how governments think about growth. It is also moving faster than almost anything institutions are used to managing, regulating or teaching. borrowing a case study or a guest lecture. That division of labour made sense when AI was a niche capability. It does not make sense now that AI is a general‑purpose force reshaping industries, businesses, competition and society all at once. If our teaching stays siloed, we will continue to produce graduates who can build systems but not question them, or who can question systems but have no real grip on how they work. Neither is good enough for what employers, companies, regulators and governments need from the next generation of leaders.
Business Impact • ISSUE 5 • 2026
11
Made with FlippingBook - Share PDF online