Artificial intelligence is the fastest-growing area of graduate study in the United States, and US universities offer more AI and machine learning master’s degrees than anywhere else in the world. That choice is great news, but it also makes it hard to know where to start. In this guide we look at ten of the strongest AI-related master’s programmes in the USA: what each one teaches, how long it takes, what it costs, and who it suits best.
How we chose them: every programme on this list is taught on campus (so international students can apply for an F-1 student visa), has a curriculum built around AI or machine learning, and comes from a university with a strong record in AI research. They are not ranked against each other: the “best” one is the one that fits your background, budget and career goals.
At a glance
| University | Programme | Typical length | Best for |
|---|---|---|---|
| Carnegie Mellon University | MS in Machine Learning | 16 months | Deep technical and research skills |
| Carnegie Mellon University | MS in AI and Innovation | 4 semesters | Building AI products and start-ups |
| Stanford University | MS in Computer Science (AI specialisation) | About 6 quarters | Silicon Valley careers |
| Northwestern University | MS in Artificial Intelligence | Autumn to the following spring | Small classes, industry focus |
| Duke University | MEng in AI for Product Innovation | 12 or 16 months | A fast, practical degree |
| Columbia University | MS in Computer Science (Machine Learning track) | 30 points | New York and finance or media jobs |
| Georgia Tech | MS in Computer Science (Machine Learning specialisation) | 30 credit hours | Strong value from a top public university |
| Johns Hopkins University | MSE in Computer Science (AI track) | About 3 semesters | AI in health and science |
| University of Southern California | MS in Computer Science – Artificial Intelligence | 32 units | Los Angeles tech and entertainment |
| University of Illinois Urbana-Champaign | Master of Computer Science (AI track) | As little as 3 semesters | A coursework-only route |
1. Carnegie Mellon University: MS in Machine Learning
Carnegie Mellon (CMU) in Pittsburgh has the world’s first university department dedicated entirely to machine learning, and its master’s is one of the most demanding AI degrees anywhere. It is a 16-month programme, usually taken over three semesters, with a summer practicum (an internship or research project in machine learning) after the first year.
- What you study: six core courses (machine learning, statistics, probabilistic graphical models, optimisation, machine learning in practice, and deep learning or deep reinforcement learning) plus three electives.
- Entry: very competitive, with a strong maths and programming background expected. CMU applications open once a year for an August start, and the GRE was optional for the 2026 intake.
- Cost: CMU’s School of Computer Science master’s rate for 2026–27 is US$31,100 per semester. The department says the programme does not offer financial support.
- Best for: students with a strong maths background who want to become machine learning engineers or researchers, or go on to a PhD.
2. Carnegie Mellon University: MS in Artificial Intelligence and Innovation
CMU’s second entry is very different. The MS in AI and Innovation (MSAII), run by the Language Technologies Institute, combines AI with entrepreneurship. Over four semesters you take courses in AI, machine learning, deep learning, natural language processing and generative AI, then work in a team to turn an idea into a working product.
- Highlights: a summer industry internship between the first and second year, and a capstone project that aims to produce a product ready for a company or for start-up investment.
- Best for: students who want to launch AI products or start their own company, rather than focus on research.
3. Stanford University: MS in Computer Science (Artificial Intelligence)
Stanford sits at the heart of Silicon Valley, and many of the ideas behind modern AI came from its labs. Its MS in Computer Science lets you specialise in artificial intelligence, choosing from advanced courses in machine learning, natural language processing, computer vision and robotics.
- Length: 45 units. Stanford says most students finish in about six quarters (roughly two academic years).
- Cost: 2026–27 graduate engineering tuition is US$23,239 per quarter for a full load (11–18 units).
- Best for: students with a strong computer science background who want close links with leading tech companies and start-ups.
4. Northwestern University: MS in Artificial Intelligence
Northwestern’s MSAI, near Chicago, is one of the few US master’s degrees named simply “Artificial Intelligence”. It is deliberately small, with about 45 students a year, and it focuses on designing AI systems and putting them to work in real businesses.
- Two tracks: the traditional MSAI for students heading into industry, and MSAI+X for people from other fields who want to use AI in their own area. MSAI+X starts with a summer bootcamp in programming and maths.
- Length: from the first autumn to the end of the following spring quarter.
- Cost: US$23,662 per quarter for 2026–27. Northwestern puts the total cost, including living costs, at about US$125,000, and says the programme offers no scholarships or teaching assistantships.
- Best for: students who want small classes, lots of contact with industry, and a clear focus on applied AI.
5. Duke University: Master of Engineering in AI for Product Innovation
Duke’s MEng is a practical, career-focused degree that you can finish in 12 months (or 16 months if you want more electives). Alongside technical AI and machine learning courses, you take product development courses created with Duke’s law and business schools, covering the business, legal and ethical sides of building AI products.
- Highlights: an online Python and maths bootcamp before you start, an industry project for a sponsoring organisation, a summer internship or industry project, and a capstone.
- Cost: Duke estimates total tuition of US$82,064 for the 12-month on-campus route in 2026–27.
- Best for: students who want a short, hands-on degree that leads straight into an industry job.
6. Columbia University: MS in Computer Science (Machine Learning track)
Columbia, in New York City, offers a dedicated Machine Learning track within its MS in Computer Science. The degree is 30 points, all at advanced level, with at least 6 points from Columbia’s most advanced (6000-level) courses.
- What you study: fundamental courses such as machine learning and computational learning theory, plus options in natural language processing, artificial intelligence, robotics and more.
- Best for: students who want to work in New York’s finance, media and tech sectors, all of which hire heavily in AI.
7. Georgia Tech: MS in Computer Science (Machine Learning specialisation)
The Georgia Institute of Technology in Atlanta is one of the top public universities for computing in the USA. Its MS in Computer Science is 30 credit hours, and machine learning is one of its specialisations, covering topics such as deep learning and reinforcement learning.
- Value: as a public university, Georgia Tech is often cheaper than private universities of a similar standard. Check the Bursar’s office for the current international rate.
- Best for: students who want a top-ranked technical degree at a lower cost than most private universities.
8. Johns Hopkins University: MSE in Computer Science (Artificial Intelligence track)
Johns Hopkins in Baltimore is famous for medicine and science, which makes its AI track a strong choice if you want to apply AI to health, biology or other research fields. Students on the track can use the resources of the university’s Data Science and AI Institute.
- Structure: eight approved courses plus a research project, or ten courses. You can choose a coursework-only option or combine it with a project, essay or thesis.
- Length: Johns Hopkins says students take about three semesters on average.
- Cost: full-time graduate tuition was US$66,670 a year in 2025–26. Check the university’s site for the 2026–27 rate.
- Best for: students interested in AI for healthcare, life sciences or research.
9. University of Southern California: MS in Computer Science – Artificial Intelligence
USC’s Viterbi School of Engineering in Los Angeles offers a dedicated MS in Computer Science with an artificial intelligence focus. It is 32 units of coursework, with electives such as advanced natural language processing.
- Research: there is no thesis option, but after your first semester you can ask to do directed research with a faculty member.
- Entry: USC says the GRE is not required for 2027 applications.
- Best for: students who want to study in Los Angeles, with access to its tech, gaming and entertainment industries.
10. University of Illinois Urbana-Champaign: Master of Computer Science (AI track)
The University of Illinois Urbana-Champaign has one of the largest and most respected computer science departments in the USA. Its on-campus Master of Computer Science is a coursework-only degree of 32 credit hours that you can finish in as little as three semesters. You can choose one of eight tracks, including artificial intelligence and data science.
- Best for: students who want a respected degree focused on taught courses rather than research, finished in a short time.
Studying online instead
Several top universities also offer AI master’s degrees fully online. They cost much less, but an online degree does not qualify you for a US student visa or for work in the USA after graduation, so they suit people who want to stay in their home country and job.
- University of Texas at Austin, MS in Artificial Intelligence (online): US$10,000 tuition (plus fees), taken over 18–36 months.
- Georgia Tech, Online MS in Computer Science (OMSCS): the same 30-hour degree as on campus, with a machine learning specialisation, at a small fraction of the on-campus cost.
- University of Pennsylvania, MSE in Artificial Intelligence (online): the first AI master’s of its kind in the Ivy League. No GRE is required, and you can finish in as few as 16 months.
- University of Illinois, Online Master of Computer Science: delivered through Coursera, with artificial intelligence courses available.
How to choose the right AI master’s for you
- Your background: most programmes expect a degree in computer science, engineering, maths or a related subject, with solid programming and maths (linear algebra, calculus, probability and statistics). If you’re coming from another field, look at options with a bootcamp, such as Northwestern’s MSAI+X or Duke.
- Research or industry: if you might want a PhD, choose a programme with research or thesis options (such as CMU’s MS in Machine Learning or Johns Hopkins). If you want a job quickly, practical degrees with internships and industry projects (Duke, Northwestern, CMU’s MSAII) are a better fit.
- Total cost: remember to add living costs to tuition. Big cities such as New York, Los Angeles and the San Francisco Bay Area are much more expensive than Pittsburgh, Atlanta or Urbana-Champaign.
- Length: one-year programmes cost less overall, but longer ones leave room for an internship, which often leads to a job offer.
- Working after you graduate: on an F-1 visa you can usually work for up to 12 months after graduating under Optional Practical Training (OPT). Graduates of STEM-designated degrees can apply for a 24-month extension. Most computer science and AI master’s are STEM-designated, but check with each university.
For more about studying in the US, including living costs, visa requirements and post-study work rights, see our Study in the United States guide, or browse US institutions on our Universities page.
Fees, course details and entry requirements were checked on each university’s official website in October 2026. They change every year, so always confirm the latest information with the university before you apply.