Harvard University’s Professional and Lifelong Learning division currently lists six free courses in artificial intelligence, machine learning, and programming on its catalog. The courses span beginner through intermediate levels and are delivered entirely online through edX and Harvard’s OpenCourseWare (OCW) platform. Learners can audit all six at no cost. Optional verified certificates on edX carry fees ranging from $149 to $299, depending on the course.

The offerings run the gamut from a widely recognized computer science introduction designed for non-specialists, to hands-on technical courses in Python-based machine learning and embedded AI systems. Three of the six form a dedicated certificate sequence in Tiny Machine Learning, co-taught by a Harvard School of Engineering faculty member and a senior AI advocate at Google.

“Learners can audit all six courses at no cost. Optional verified certificates carry fees ranging from $149 to $299.”

The free audit model means anyone with an internet connection can work through lecture content, readings, and in many cases problem sets, without paying. Verified certificates, which involve identity checks through edX and are generally regarded as more credential-worthy for professional use, require a fee.

Below is a course-by-course breakdown based on Harvard’s official course listings.

The Six Courses

Course 01

CS50’s Introduction to Artificial Intelligence with Python

Intermediate 7 weeks  |  10–30 hrs/week  |  Self-paced Instructors: David J. Malan & Brian Yu

This course covers the concepts and algorithms underpinning modern AI, including graph search, classification, optimization, reinforcement learning, and neural networks, through hands-on Python projects. By the end, learners have experience with machine learning libraries and enough grounding to design basic intelligent systems. It is part of Harvard’s Professional Certificate in Computer Science for Artificial Intelligence on edX and is taught by Gordon McKay Professor David J. Malan and Senior Preceptor Brian Yu of Harvard’s School of Engineering and Applied Sciences.

View course on Harvard PLL →
Course 02

Machine Learning and AI with Python

Intermediate 6 weeks  |  4–5 hrs/week  |  Self-paced Harvard School of Engineering and Applied Sciences

Using real-world datasets, this course builds learners’ understanding of decision trees as a foundation, then extends into bagging, random forests, and gradient boosting. The curriculum emphasizes working with Python to detect overfitting, apply cross-validation, and avoid biased outcomes. It is aimed at learners who already have some Python exposure and want to move into practical data science work.

View course on Harvard PLL →
Course 03

Data Science: Building Machine Learning Models

Introductory 8 weeks  |  2–4 hrs/week  |  Self-paced Instructor: Rafael Irizarry, Professor of Biostatistics, Harvard T.H. Chan School of Public Health

Part of Harvard’s Professional Certificate in Data Science, this course teaches popular machine learning algorithms, principal component analysis, regularization, and cross-validation through the lens of building a movie recommendation system. The practical framing makes abstract concepts such as training sets and overtraining more concrete. Learners who complete this course gain a solid foundation in the methodology behind widely used recommendation and prediction products.

View course on Harvard PLL →
Course 04

Fundamentals of TinyML

Introductory Self-paced  |  2–4 hrs/week Instructors: Vijay Janapa Reddi (Harvard SEAS) & Laurence Moroney (Google)

The first course in Harvard’s TinyML Professional Certificate series introduces learners to machine learning at the edge, meaning on small, resource-constrained devices such as smartphones and microcontrollers. Topics include data collection for ML, training and deploying basic models, fundamentals of deep learning, and responsible AI design. No prior embedded systems experience is required, making this the recommended starting point for learners new to the field.

View course on Harvard PLL →
Course 05

Applications of TinyML

Intermediate 6 weeks  |  2–4 hrs/week  |  Self-paced Instructors: Vijay Janapa Reddi (Harvard SEAS) & Laurence Moroney (Google)

The second course in the TinyML series moves from theory into real-world applications. Learners examine the code behind voice recognition features, including how devices respond to wake words such as “OK Google” and “Alexa,” as well as visual wake words, anomaly detection, and gesture recognition. The course also covers dataset engineering and responsible AI development, and draws on industry case studies to contextualize deployment challenges on deeply embedded devices.

View course on Harvard PLL →
Course 06

CS50’s Computer Science for Business Professionals

Introductory 9 weeks  |  Self-paced  |  Course began April 1, 2026 Instructor: David J. Malan

A top-down variant of Harvard’s flagship CS50x course, this program is designed for managers, product managers, founders, and enterprise decision-makers who need conceptual fluency in technology without necessarily becoming practitioners. Rather than building from low-level implementation detail, it prioritizes high-level design decisions and their business implications. Topics include cloud computing, networking, privacy, scalability, web technologies, security, artificial intelligence, and databases. A revised version of the course launched April 1, 2026. Learners can take it free through Harvard’s OCW platform or enroll on edX for a verified certificate.

View course on Harvard OCW →

Notes for Prospective Learners

On certificates: Auditing is free on edX for all six courses. Verified certificates, which carry an edX identity-verification process and are generally considered more credible for professional use, are available for an additional fee: $149 for the Data Science course, $299 for the two CS50 AI and Machine Learning with Python courses, and at varying prices for the TinyML series. The CS50 Business course offers a free certificate of completion through Harvard’s own platform upon passing all assignments.

On prerequisites: The TinyML Fundamentals and CS50 Business courses have no formal prerequisites. CS50 AI and Machine Learning with Python recommend some prior Python experience. The Data Science: Building Machine Learning Models course benefits from basic statistics knowledge and is part of a broader nine-course Data Science certificate program on edX, though individual courses can be taken independently.

On the Deploying TinyML course: A third TinyML course, Deploying TinyML, is also listed as free on Harvard’s catalog. It covers programming TensorFlow Lite for microcontrollers and hands-on deployment to physical Arduino-based hardware. It is not counted among the six in this article as it was not part of the original batch highlighted in recent coverage, but it is available through the same edX platform at no audit cost.

All six courses are accessible now. Enrollment links for each course are available through Harvard’s Professional and Lifelong Learning portal at pll.harvard.edu and, for CS50 Business, directly at cs50.harvard.edu/business.