No dedicated accreditation standard for AI degrees is in force in the United States or Canada as of October 24, 2026, so a school's place on a list of AI programs says little about quality control. ABET, the main US accreditor of computing programs, has a commission-approved but not yet finalized set of AI and machine learning program criteria that are expected to take effect in Fall 2027. Until then, an applicant has to check institutional recognition first, then any program-level accreditation that covers the related computer science or engineering degree.
Key takeaways
- ABET accredits programs only, never whole institutions, and its current 2026-2027 computing criteria have program criteria for computer science, data science, cybersecurity, information systems, and information technology, but none for AI.
- Proposed ABET criteria for AI and machine learning programs would require at least 42 semester credit hours of AI coursework; final approval was expected on November 6, 2026.
- In Canada, quality assurance of degrees is a provincial and territorial duty, and professional accreditation covers engineering (CEAB) and computer science (CIPS) separately.
- Accreditation sets a minimum standard. It does not rank programs against each other.
What accreditation does and does not certify
Accreditation answers a narrow question: does a school or program meet a published standard? It does not say which school is better. In the United States, the Department of Education does not accredit schools itself. It recognizes independent accrediting agencies, and a school that wants its students to receive federal aid generally needs accreditation from one of them. That is institutional accreditation, and it applies to the school as a whole.
Programmatic accreditation is a second, separate layer that covers one degree inside a school. ABET states that it offers accreditation "to programs only, not to degrees, departments, colleges, institutions, or individuals." A university can therefore hold institutional accreditation while only some of its computing programs carry ABET status, and a program can be well taught without ABET status at all, because the review is voluntary.
Where AI sits in ABET's current rules
ABET's Computing Accreditation Commission publishes criteria for each review cycle. The document effective for the 2026-2027 cycle includes program criteria for computer science, cybersecurity, data science and analytics, information systems, and information technology, with CSAB named as lead society. A program titled "Artificial Intelligence" is not covered by a named set of criteria in that version. Reviewers apply the general criteria and the closest program criteria instead.
The same document carries proposed criteria for programs with artificial intelligence, machine learning, or similar terms in their titles. As proposed on October 24, 2025, they would add a student outcome on applying AI theories, models, and techniques to design and implement solutions, and a curriculum with at least 42 semester credit hours of AI coursework and at least nine credit hours of mathematics and statistics. The AI coursework would have to cover foundations such as reasoning and heuristic search, machine learning including deep learning, and ethics and responsible AI, plus an application area and a major project. CSAB reported on July 27, 2026 that the commission had approved the criteria and that the Computing Area Delegation was expected to give final approval on November 6, 2026, with an effective date of Fall 2027. As of October 24, 2026, they remain proposed.
How the layers compare in the two countries
The table sets out who checks what. It is a simplified orientation based on each body's own description.
| Layer | United States | Canada |
|---|---|---|
| School or university | Independent accreditors recognized by the Department of Education | Provincial and territorial governments and their quality assurance agencies; no national network of agencies |
| Computer science and software degrees | ABET, computing criteria | CIPS Computer Science Accreditation Council, for bachelor's-level programs |
| Engineering degrees (for example computer engineering) | ABET, engineering criteria | Canadian Engineering Accreditation Board (CEAB), run by Engineers Canada |
| Dedicated AI degree | Proposed ABET AI/ML criteria, not yet final | No AI-specific accreditation named on the CIPS program page |
The Council of Ministers of Education's information service, CICIC, explains that Canadian postsecondary quality assurance rests with provincial and territorial governments, that no national network of quality assurance agencies exists, and that the Association of Accrediting Agencies of Canada is a membership network of professional accreditors, not a national regulator.
Canadian professional accreditation in detail
Engineers Canada reports 308 accredited engineering programs at 47 institutions. CEAB reviews undergraduate engineering programs only and accredits no courses, individuals, or institutions, and the status supports licensure as a professional engineer. CIPS accredits postsecondary degree and diploma programs, including bachelor's-level computer science and software engineering, but not certificates, training, or doctoral programs. Its program page does not mention artificial intelligence. A student in Canada who wants an AI-focused degree will usually find it filed under computer science, engineering, or a data science stream, so the useful check is whether that parent program holds CIPS or CEAB status.
A checklist for any program
- Is the school accredited by a recognized institutional accreditor (US) or authorized to grant degrees by its province or territory (Canada)?
- Does the department state which programmatic accreditation, if any, covers this degree, and since when?
- Is the AI content inside an accredited computer science or engineering program, or a stand-alone degree with no external review?
- Does the curriculum page list required courses in mathematics, programming, machine learning, and ethics?
- Does the accreditor's own search tool, not the school's brochure, confirm the status?
ABET publishes a searchable list of accredited programs, and CIPS publishes lists of the programs it accredits. The information here is general. Immigration, credit transfer, and licensing questions belong with the school's registrar, the relevant regulator, or an advisor.
FAQ
Does ABET accredit AI programs today?
Not under a dedicated standard. ABET accredits computing programs by general criteria plus the closest program criteria. AI and machine learning criteria were approved by the commission in July 2026 and awaited final approval, with an effective date of Fall 2027.
Is an unaccredited AI degree worthless?
No. Programmatic accreditation is voluntary, and many strong programs have none. It matters most where a profession depends on it, such as engineering licensure in Canada, so the need for it depends on the career path.
Who accredits universities in Canada?
No federal body does. Provincial and territorial governments are responsible, each with its own arrangements, and professional bodies accredit particular programs.
Does accreditation rank schools?
No. Engineers Canada says explicitly that accreditation is not a ranking, and the same logic applies to ABET and CIPS, which certify that standards are met.
Related reading
- Comparing institutions for AI studies
- Institutions offering AI and data science programs
- Impact and potential of AI in education
- How AI changes the learning experience
More in AI in education.
Last verified: October 2026
Sources
- What is accreditation, ABET, 2026
- Criteria for Accrediting Computing Programs, 2026-2027, ABET Computing Accreditation Commission, incorporating changes approved October 24, 2025
- New AI/ML program criteria expected to be approved for Fall 2027 implementation, CSAB, July 27, 2026
- Quality assurance in postsecondary education: an overview, CICIC (Council of Ministers of Education, Canada), 2026
- Accreditation, Engineers Canada, 2026
- Accredited program categories (program page), CIPS, 2026