No official source ranks institutions for artificial intelligence studies, but two public classification and outcome systems let applicants compare programs on facts. Canada and the United States share a classification code, 11.0102 Artificial intelligence, and the U.S. College Scorecard publishes earnings and debt by field of study. A list of the "best" schools would rest on someone's chosen weights, while these sources show what each number covers.
Key takeaways
- CIP Canada 2021 and the matching U.S. CIP 2020 both list 11.0102 Artificial intelligence as a program class, with machine learning and robotics programming given as examples.
- The College Scorecard groups programs by four-digit CIP code, so an AI program coded 11.0102 sits in the same group as general computer science (11.01).
- Scorecard earnings and debt describe only students who received federal Title IV aid, and debt counts only federal loans.
- Statistics Canada links student records to earnings records, but its public field-of-study groupings are broad, not program by program.
What the classification code says an AI program is
Statistics Canada and the U.S. National Center for Education Statistics revised the Classification of Instructional Programs together. The U.S. edition is CIP 2020 and the Canadian edition is CIP Canada 2021, which has 50 series, 454 subseries, and 2,119 instructional program classes. Class 11.0102, Artificial intelligence, sits under 11.01, computer and information sciences and support services, general.
Its definition covers any program that focuses on symbolic inference, representation, and simulation by computers and software of human learning and reasoning, plus computer modeling of human motor control. Listed courses include computing theory, cybernetics, human factors, and natural language processing. Neighboring classes keep related fields apart: human computer interaction is coded 30.3101, and mechatronics, robotics, and automation engineering has its own code outside computer science. A program that teaches machine learning but is filed under 11.0101 or under statistics will therefore not appear under the AI class, and the reverse can also occur.
What the College Scorecard can and cannot separate
The U.S. Department of Education publishes College Scorecard data by field of study. Institutions report completions by six-digit CIP code to the Integrated Postsecondary Education Data System, and the Scorecard then sums them to four digits. The documentation gives the reason: larger groups mean fewer values hidden for privacy. The result is that 11.0102 is merged with 11.0101 (general), 11.0103 (information technology), 11.0104 (informatics), and the other 11.01 classes.
The unit of analysis is an institution, a four-digit code, and a credential level. Earnings come from tax records (W-2 forms and self-employment income) and are reported at several intervals after completion. The October 2026 documentation lists measures at years one, four, five, and seven. Where a school's own count is too small, the data files carry a "PS" code for privacy suppression, and national figures by code and credential level are provided as a fallback.
Whose outcomes are counted
Scorecard earnings and cumulative debt are derived from the National Student Loan Data System, so they describe only students who received federal Title IV grants or loans. The documentation tells users not to assume the values represent all graduates of a field. Debt covers federal loans for borrowers only, excludes private and state loans, and leaves out capitalized interest. A school with many students who pay without aid can therefore look very different from the numbers shown.
The Canadian side: broad groupings, not named programs
Statistics Canada's longitudinal platform for education and labor-market data (ELMLP) links Postsecondary Student Information System records with administrative earnings data. Its overview, issued August 15, 2024, describes median employment income for graduates by qualification, field of study, age group, and gender, with related tables measured two, five, or ten years after graduation. The fields are shown as primary groupings and as STEM versus BHASE (non-STEM) groupings. Nothing in that overview describes a separate line for a six-digit class such as 11.0102, so Canadian applicants should treat these tables as field-level context, not as a program comparison.
A comparison sheet built from these sources
| Question | Where the answer is found | Limit to keep in mind |
|---|---|---|
| How is the program classified? | Registrar or program office (ask for the CIP code); CIP Canada 2021 or CIP 2020 for the definition | Titles and codes do not always match |
| What do graduates earn in the U.S.? | College Scorecard field-of-study files | Four-digit group, federally aided students only |
| How much debt do they carry? | Same files | Federal loans only, interest excluded |
| What do graduates earn in Canada? | Statistics Canada ELMLP tables | Broad field groupings, public postsecondary graduates |
| What is actually taught? | Required-course list on the program page | Only the school can supply it |
Checks before treating any list as a ranking
- Does the list say which measures it uses and how much each counts?
- Does the program appear under the AI class or under a general computer science class?
- Are earnings and debt figures from the same group of students?
- Is the credential level (bachelor's, master's, certificate) the one being compared?
- Is the data year stated, and is it recent enough to matter for a new applicant?
FAQ
Is there an official ranking of AI programs in Canada or the United States?
None was found in the classification or outcome systems described here. The CIP code only names a category of program, and the Scorecard and Statistics Canada tables report outcomes without ordering schools by quality.
Can the College Scorecard show earnings for an AI degree alone?
Not by the six-digit class. The Scorecard uses four-digit codes, so an AI program is grouped with other 11.01 programs, and small groups may be suppressed.
Why might a program teaching machine learning not be coded as artificial intelligence?
Schools assign codes themselves. Machine learning is listed as an example under 11.0102, but a program can sit in computer science, data science, or statistics, depending on the department that offers it.
Does this page advise which school to choose?
No. It explains what public data measure. Admission requirements, cost, immigration rules for international students, and fit with personal goals need to be checked with each institution.
Related reading
- Artificial intelligence schools and research institutions
- Institutions offering AI and data science programs
- How AI is affecting education
- Limits of AI in education
More in AI in education.
Last verified: October 2026
Sources
- Technical Documentation: College Scorecard Data by Field of Study, U.S. Department of Education, version October 2026 (source page)
- Classification of Instructional Programs (CIP) Canada 2021 Version 1.0, Statistics Canada, product 12-590-X, 2022
- Introduction to the Classification of Instructional Programs (CIP) Canada 2021 Version 1.0, Statistics Canada, 2022 (source page)
- Statistics Canada, ELMLP overview and associated datasets, issued August 15, 2024