Universities are taking different routes to prepare students for AI: some are weaving it into existing subjects, while others are changing what and how they teach. Nanyang Technological University plans to include AI in 40% of courses across all 52 undergraduate programmes by 2030, a shift that puts AI literacy well beyond the computer science department.
Watch Desk analysis
What happened
The Korea Times reports that NTU’s current figure is 5%. Half of its planned AI-integrated courses will use AI to personalise learning; the other half will teach students to develop and manage AI agents. Students are also expected to work with agents on industry projects.
Other institutions are changing the rules around learning and assessment. The National University of Singapore has made quantitative reasoning and computational thinking compulsory for undergraduates, and is putting more emphasis on students’ reasoning than on answers alone. King’s College London is exploring AI-informed assessment and gives law students access to professional software used by law firms, so they can practise evaluating its output.
At the London School of Economics, students have access to Claude and staff to Microsoft Copilot, while faculty members are given time to redesign courses around AI. South Korea, by contrast, has added more than 4,000 advanced-technology seats since 2024, but limited further additions to 434 for 2027 amid concerns about shortages of staff and facilities.
Why it matters
The question is not simply how many graduates can use AI tools. Universities are deciding whether to train more specialists, teach AI across existing disciplines, or change assessment so students must show how they reached an answer. Those choices will shape what graduates can do at work, and whether they can judge when a fluent machine answer needs checking.
The contrast is especially useful for South Korea, where rapid expansion of specialist places is meeting practical limits. The Korea Times says applications to AI-related departments at 20 major universities rose 16% this year, but more seats alone may not teach students in other fields how to work with AI.
Our read
The strongest idea here is to treat AI as something students must learn to use and question within their own subject, rather than a reason to turn every student into an aspiring engineer. NTU’s plans are ambitious; NUS’s focus on reasoning is a useful reminder that assessment still needs to test the student, not merely the quality of the machine’s answer.
What to watch
- Whether NTU reaches its 40% target by 2030, and what those courses actually teach.
- Whether universities publish evidence that redesigned assessments measure students’ reasoning.
- Whether South Korea’s expansion of AI programmes addresses shortages in teaching staff and facilities.
Discussion spark: Should universities prioritise AI skills across every subject, or focus limited teaching capacity on training more specialists?
Sources and evidence
- Source update (18 October 2026, 04:16 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.