Five things the Provost should know
- AI is reaching every profession, but at USEK it sits almost only in computing.29 of 2,500 course descriptions name AI; 14 of them in Arts and Sciences, home of computer science. Law, Theology and Medicine have 0, 0 and 1.
- The frontier moves in months, not catalogue cycles.The length of task the best AI model completes on its own (at a 50% success rate) went from 3.4 to 17.4 hours in 8 months. 49 of 135 DeepLearning.AI courses were updated in the last 90 days.
- Law: no USEK course yet asks how law treats AI.None of 179 law and political science descriptions name AI. Harvard Law has at least 14 such courses this year, including one where students build and work with AI agents. Argentina has proposed companies run by AI, with human shareholders optional.
- Business: AI appears in finance electives, not in how firms are run.2 of 155 business descriptions name AI. None covers AI agents in operations, AI-driven marketing or AI governance. MIT Sloan names AI in 44 of 302 subjects.
- No faculty teaches students to think about the future.0 of 2,500 descriptions name foresight, futures thinking or scenario planning, and the rubric finds no course that teaches futures methods. Leading peers teach it across disciplines. A General Education course built on UNCERTAIN's nine steps would reach every student.
Where each faculty stands
Number of course descriptions in each faculty that name the theme. Click a faculty to zoom in. A dot means the theme has not been read for that faculty yet.
| Theme | Arts & Sciences | Engineering | Business | Architecture & Design | Medicine | Law & Political Sc. | Theology | General Ed. | Where leading peers are |
|---|---|---|---|---|---|---|---|---|---|
| Names AI at all | 14 | 5 | 2 | 5 | 1 | 0 | 0 | 2 | All 9 leading universities tracked name machine learning; generative AI went from 2 to 8 of 9 since 2021. |
| Frontier AI: agents, generative AI, reasoning, AI-assisted work | 1 | 1 | 0 | 0 | 0 | 0 | 0 | 1 | AI agents went from 0 to 3 of 9 peers since 2021; AI-assisted software work from 0 to 3. |
| Responsible AI and technology ethics | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 3 to 6 of 9 peers since 2021. |
| Converging technology: quantum, robotics, IoT, blockchain, XR, green | 13 | 13 | 1 | 4 | 0 | 0 | 0 | 0 | Quantum computing went from 5 to 9 of 9 peers; blockchain from 3 to 6. |
| Law and governance of AI: liability, regulation, legal status | · | · | 0 | · | · | 0 | · | · | Harvard Law has at least 14 courses this year whose descriptions name AI, from liability for generative AI advice to frontier AI governance. |
| Business with AI agents and AI-driven marketing | · | · | 0 | · | · | 0 | · | · | MIT Sloan: 44 of 302 subjects name AI, including a Generative AI Lab that teaches agentic workflows. |
| Foresight and futures thinking | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | Stanford and NUS teach it across engineering, business, policy and design; Edinburgh has a Futures Institute. |
| Course descriptions read | 984 | 588 | 155 | 238 | 157 | 179 | 145 | 54 |
How deep each faculty goes
Every course description scored on the Course Frontier Rubric: does the course help students understand AI, apply it, or create with it, and does it explore or shape futures? Of 2,500 courses, 29 reach any AI level, 18 reach Apply and 1 reaches Create. Only 1 course reaches the futures layer: PSY300, which aims at a desired personal future without teaching futures methods. With their briefs, 6 business and law courses move to Explore.
| Faculty | Courses | AI: Understand | AI: Apply | AI: Create | Futures: Explore or Shape |
|---|---|---|---|---|---|
| Arts & Sciences | 984 | 5 | 7 | 1 | 1 |
| Engineering | 588 | 1 | 8 | 0 | 0 |
| Business | 155 | 0 | 2 | 0 | 0 |
| Architecture & Design | 238 | 2 | 0 | 0 | 0 |
| Medicine & Medical Sciences | 157 | 1 | 0 | 0 | 0 |
| Law & Political Sciences | 179 | 0 | 0 | 0 | 0 |
| Theology | 145 | 0 | 0 | 0 | 0 |
| General Education | 54 | 1 | 1 | 0 | 0 |
Levels are proposed by rules from each course description and are confirmed by faculty on the platform. The rules were tested on 20 labelled course texts (98 of 100 scores exact); five of those texts were added after a first catalogue run surfaced false matches, and the rules were fixed until they scored zero.
What the outside world is doing
of work the best model completes on its own at a 50% success rate, from 3.4 hours 8 months earlier.
of 1,001 technical postings at six labs ask for agents and tool use; 44% for comfort with ambiguity.
of 1,356 technical postings at four leading AI firms ask for cloud and DevOps; 27% for generative AI.
Jobs at the top 30 to 50 employers of USEK graduates, graduates' new titles and new academic posts, from LinkedIn. Not measured yet.
Zoom into a faculty
Business and Law are open: they were read in depth this week against MIT Sloan and Harvard Law. Each quote is taken word for word from the catalogue or course page.
Arts & Sciences984 courses · 14 name AI · 13 of 18 technology areas · 0 foresight · Read in depth
CSC, EDU, THT, PSY, CRCS, CHM. Closest AI benchmark: Fixing real software bugs, best score 0.83.
Computing (Computer Science) sits here and is read in depth: Skills Frontier for skills and jobs, and Course Briefs for what to keep, compress and add, course by course, with this week's MOOC and lab modules.
Engineering588 courses · 5 name AI · 10 of 18 technology areas · 0 foresight · Catalogue counts only
GMC, GCH, GEL, GBM, GCV, GAG. Closest AI benchmark: CAD models from a description, best score 0.74.
Read from the catalogue only so far. Peer, MOOC, job and course-level reading for this faculty comes next.
Business155 courses · 2 name AI · 3 of 18 technology areas · 0 foresight · Read in depth
LCB, MKT, BUS, MGT, FIN, ACT. Closest AI benchmark: Professional deliverables against industry experts, best score 0.50.
What USEK business courses teach now
2 of 155 course descriptions name AI, both finance electives (FIM640 Computational Finance, FIM655 Artificial Intelligence for Finance). FIM640: This course will cover the major aspects of Data Science and Artificial Intelligence.
Next to AI: fintech and blockchain, digital marketing, digital transformation and business analytics. FIN410 covers how financial technology (fintech) is impacting areas within the investment industry, such as investment analysis, automated advice, and risk management.
- BUS325 Blockchain & Cryptocurrency
- BUS335 Fintech
- FIN410 Investment Management
- MKT350 Digital Marketing
- MKT685 Digital Marketing Strategy
- AUD650 Information Systems and Controls (ISC) – CPA
- BUS440 Digital Transformation 3cr.
- HRM565 Mastering the HR Tools
- FIM640 Computational Finance
- ITB413 Business Data Analytics
- MGT335 Technology and Innovation Management
No business course names AI agents, AI in marketing, AI-native ventures, the law of AI, or AI ethics. MGT335 studies start-ups, not AI-native ones: The course will focus on entrepreneurial firms (start-ups and established firms) to analyze success and failure cases of innovation.
What MIT Sloan teaches
44 of 302 subjects in its 2026-27 catalogue name AI.
- 15.630 AI Strategy, Governance, and Ethics.
Examines how artificial intelligence is reshaping organizational strategy, governance, and ethical responsibility.
- 15.566 Generative AI Lab. Tools include
AI evals, RAG/vector databases, agentic workflows and frameworks, prompt engineering
. - 15.814 Marketing Innovation.
Subject updated continually to highlight the latest use of artificial intelligence for marketing.
- 15.622 The Law of Artificial Intelligence and Big Data.
topics generally include the legal frameworks of intellectual property rights, cybersecurity, privacy, bias, fintech, fair competition and fair dealing in digital markets
. - 15.562 Web3 and Strategy. Covers
decentralized finance (DeFi) and organizations (DAOs), smart contracts
. - 15.679 Frontline Lab.
optimizing human-AI collaboration while developing solutions that balance job quality with competitive advantage.
Architecture & Design238 courses · 5 name AI · 4 of 18 technology areas · 0 foresight · Catalogue counts only
ARCH, CVA, DAA, DIM, ADI, CTA. Closest AI benchmark: CAD models from a description, best score 0.74.
Read from the catalogue only so far. Peer, MOOC, job and course-level reading for this faculty comes next.
Medicine & Medical Sciences157 courses · 1 name AI · 0 of 18 technology areas · 0 foresight · Catalogue counts only
MEDR, SIN, MED, FHS, MEDF, BIO. Closest AI benchmark: Expert-level science questions, best score 0.96.
Read from the catalogue only so far. Peer, MOOC, job and course-level reading for this faculty comes next.
Law & Political Sciences179 courses · 0 name AI · 0 of 18 technology areas · 0 foresight · Read in depth
DRG, ICL, DRL, RIN, CRM, POL. Closest AI benchmark: Professional deliverables against industry experts, best score 0.50.
What USEK law courses teach now
None of 179 course descriptions name AI or algorithms. The closest are two internet law courses and corporate law.
- DRG320.1 Computer and Internet Law.
This course deals with learning the essentials of computer and network technologies, and how those technologies are challenging settled legal understandings.
- ICL585 Internet Law.
The questions addressed may include issues of copyright, trademark, defamation, the Communications Decency Act, cybercrime, contracts, privacy and personal jurisdiction.
- DRG510 Advanced Corporate Law covers
the rules of powers and responsibilities applicable to each corporate player, legal person and corporate organs.
A natural home for the question of companies run by software.
What Harvard Law teaches this year
At least 14 courses in the 2026-27 catalogue name AI.
- Agentic Legal Engineering.
This hands-on seminar allows students to create and work directly with agents, discovering areas of untapped potential and unexpected danger.
- Algorithms, Rights, and Responsibilities. Covers
liability for incorrect or unlawful advice given by generative AI
. - Artificial Intelligence, Law and Healthcare.
And as AI moves beyond decision support toward increasingly autonomous agents, are our existing legal and ethical frameworks still fit for purpose?
- Debates on Frontier AI Governance.
This course will explore some of the most fundamental debates in the legal and corporate governance of frontier AI models and systems.
- Hard Cases. Asks which matters
may be exemplars of matters requiring human judgment rather than Artificial Intelligence tools; if so, why?
The live case: companies without people
Argentina's government has proposed a corporations law with a new category, the non-human corporation: an entity run by AI agents or robots that can sign contracts and hold assets on its own, with human shareholders optional rather than required.
President Milei argues that creating a legal framework for what he calls “non-human corporations” would make them easier
to regulate. It is a proposal, not yet law. No course read, at USEK or Harvard, names AI legal personhood or AI-run companies.
Theology145 courses · 0 name AI · 0 of 18 technology areas · 0 foresight · Catalogue counts only
THEO, MUSC, MUI, MTH, MUG, MSS.
Read from the catalogue only so far. Peer, MOOC, job and course-level reading for this faculty comes next.
General Education54 courses · 2 name AI · 1 of 18 technology areas · 0 foresight · Catalogue counts only
GEHS, GEAD, GEPS, GESH, GEHU, GEMG.
Read from the catalogue only so far. Peer, MOOC, job and course-level reading for this faculty comes next.
Every technology area, faculty by faculty and level
Courses naming each of 18 technology areas. Hover a cell for introductory, upper and graduate levels. Last row: share of leading-firm technical postings asking for the area.
| Faculty | AI | Computing | Frontier technology | |||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| AI agents | Generative AI and LLMs | Reasoning models | Machine and deep learning | Computer vision | Responsible AI and tech ethics | AI-assisted software work | Cybersecurity | Cloud and DevOps | Data engineering and big data | GPUs and high-performance computing | Human-computer interaction and UX | Quantum computing | Robotics and autonomous systems | IoT, edge and embedded | Blockchain and Web3 | AR, VR and spatial computing | Green and sustainable computing | |
| Arts & Sciences | 1 | 9 | 2 | 1 | 11 | 10 | 5 | 1 | 4 | 2 | 4 | 3 | 5 | |||||
| Engineering | 1 | 6 | 6 | 1 | 3 | 1 | 11 | 2 | 1 | 1 | ||||||||
| Business | 2 | 1 | 1 | |||||||||||||||
| Architecture & Design | 6 | 1 | 1 | 3 | ||||||||||||||
| Medicine & Medical Sciences | ||||||||||||||||||
| Law & Political Sciences | ||||||||||||||||||
| Theology | ||||||||||||||||||
| General Education | 1 | |||||||||||||||||
| Leading-firm demand | 22% | 27% | 0% | 19% | 0% | 5% | 5% | 6% | 38% | 30% | 10% | 15% | 0% | 2% | 1% | 0% | 0% | 0% |
Course by course
For deans and programme heads. Each brief says what to keep, compress and add in one course, with learning outcomes, assignments, assessment, the evidence behind it, and this week's MOOC and lab modules.
- Computing course briefs: 17 of 91 courses so far.
- Business and law course briefs: FIM655, MKT685 and MGT335 in Business; DRG320.1, ICL585 and DRG510 in Law.