Updated November 2026
TL;DR: Stanford, MIT, and Carnegie Mellon run summer AI programs for high schoolers, from one-week bootcamps to four-week research programs. The best university AI program for you mostly comes down to your grade, how much you can code, and where you live. Most are competitive. Several deadlines land on Feb 1, 2027, so it's worth starting this winter. And if you'd like to start an AI project before summer, you can work on one online with a mentor at Veritas AI.
If you're a high schooler who enjoys coding, a university program is a great way to try AI research. University AI programs let you learn from a university's own researchers and then build an AI project with a small team. Every program here is run by the university itself.
How to choose a university AI research program
Most university AI programs take only certain grades, so your grade is the first thing to check. Here's a quick test, based on your grade this school year.
9th grader: Stanford AI4ALL is built just for you.
8th to 11th grader who codes well: try the Summer Research Internship at Stanford's Center for Artificial Intelligence in Medicine and Imaging (AIMI).
8th to 11th grader who's new to coding: look at the Stanford AIMI Summer Health AI Bootcamp.
9th to 11th grader who can get to the Boston area: the bootcamp at the MIT Jameel Clinic, MIT's center for AI in health, is a good fit.
Junior in the US: apply to MIT's Beaver Works Summer Institute, and to Carnegie Mellon's AI Scholars if you're a US citizen or permanent resident.
Student who'd rather take an AI class than do a project: compare the options in our guide to AI programs for high school students.
For most students who can code, our pick is the Stanford AIMI Summer Research Internship. It's online, and it takes four grades and students from around the world.
University AI research programs for high school students in 2027
Six university-run AI programs at Stanford, MIT, and Carnegie Mellon take high schoolers for summer 2027. Here are the key facts as of November 2026, leaving out AI classes that have no research project.
Stanford AIMI Summer Research Internship
Who it's for: Students entering grades 9 to 12 in fall 2027 who are over 14 when their session starts. You need strong coding skills, and students outside the US can apply.
Format: Two weeks online, weekday mornings Pacific time. Stanford researchers and doctors lecture on machine learning and on AI that reads medical scans like X-rays. Then your small team does a health AI research project with Stanford mentors.
Dates: One session in mid-June and one in mid-July. Applications open in November and close in early February.
Cost: $3,000 for 2027. Need-based aid is open to US residents.
How to apply: Apply online with one recommendation.
Best for: Confident coders who are curious about AI in medicine.
Watch out for: Stanford plans to take only about 100 students.
Stanford AIMI Summer Health AI Bootcamp
Who it's for: The same grades and ages as the internship, with no coding or AI experience needed.
Format: Two weeks online, on the same schedule. You'll go to lectures, work through guided exercises and case studies, and present a team project.
Dates: The same sessions and deadlines as the internship.
Cost: $2,500 for 2027. Need-based aid is open to US residents.
How to apply: Apply online, with no recommendation letters. You can apply to both AIMI programs, but you can only attend one.
Best for: Beginners who want a first taste of health AI research.
Watch out for: The projects are guided, so you'll have less freedom than in the internship.
Who it's for: 9th graders aged 13 to 15, with no AI or coding experience needed.
Format: Two weeks, online or living at Stanford. After lectures from Stanford faculty, you'll join a small research group led by graduate students. You'll rank four topics: computer vision, medical AI, robotics, and natural language processing (how computers understand language).
Dates: Stanford hasn't posted 2027 session dates yet. Its one admissions round usually closes in early February.
Cost: Stanford lists $4,120 for the online program and $9,800 for the residential one, as of November 2026 (AI4ALL tuition page). Need-based aid is open to US and international students.
How to apply: Send transcripts and a recommendation from a math or science teacher. Our guide to Stanford's AI and CS summer programs covers each option.
Best for: 9th graders who want their first AI research project.
Watch out for: You can only apply in 9th grade, so you get one try.
Who it's for: 11th graders who are US citizens or permanent residents and will be 16 when it starts.
Format: Four weeks on Carnegie Mellon's campus in Pittsburgh. You'll take college-level courses, visit tech companies, and present a group project. You'll take an online Python course a month before, so you don't need to know how to code.
Dates: June 26 to July 24, 2027. Applications usually close in early February.
Cost: Free. Carnegie Mellon covers tuition, housing, meals, and some trips.
How to apply: Send a transcript, two recommendations, a short essay, and documents that show your family's financial need.
Best for: Juniors with financial need who want a free month of AI on a college campus.
Watch out for: It's only for US citizens and permanent residents. You also can't do another summer program at the same time.
MIT Beaver Works Summer Institute (BWSI)
Who it's for: High-achieving juniors who live or go to school in the US.
Format: A four-week summer project program, after an online course in the spring. BWSI's AI courses, including CogWorks and Medlytics, meet on Zoom. They teach machine learning for sound, images, language, and medicine. The in-person courses at MIT include self-driving race cars, drones, and underwater robots.
Dates: The online course usually opens in early February, and registered students get the summer application in March.
Cost: Free if your family's income and assets are under $200,000, and $3,000 if they're higher.
How to apply: Register, then finish enough of the online course. After that, you'll send three short essays and a recommendation (BWSI application steps).
Best for: Juniors who want to build a team AI project online, or a robotics project at MIT.
Watch out for: You must finish the online course first, and MIT doesn't provide housing for its in-person courses.
MIT Jameel Clinic AI & Health Summer High School Bootcamp
Who it's for: Students entering 10th, 11th, or 12th grade in fall 2027.
Format: One week in person at MIT, at least 7 hours a day. You'll take short courses in machine learning for health, Python, clinical AI, and drug discovery, then present a group project.
Dates: One week in the first half of July, with exact dates still to come. Early action, an earlier deadline with earlier decisions, usually closes in early November. Regular applications close in late February.
Cost: $2,000 for 2027, and scholarships can cover tuition for US students.
How to apply: Apply online. Recommendation letters are optional.
Best for: Students in the Boston area who want a short, busy start in AI for health.
Watch out for: There's no housing. It's also a week of short courses, so expect an introduction more than a research project.
How selective are university AI research programs?
In our view, Carnegie Mellon's AI Scholars, MIT's BWSI, and the Stanford AIMI internship are the hardest of these programs to get into. None of them publishes an acceptance rate. AI Scholars is merit-based and fully funded, and BWSI describes itself as a challenging program for high-achieving students. Stanford AIMI calls its internship competitive.
We think the AIMI bootcamp and the Jameel Clinic bootcamp are the easiest places to start if you're newer to coding. The AIMI bootcamp needs no coding experience and no letters. The Jameel Clinic teaches Python during the week and makes letters optional.
When are the 2027 deadlines for university AI research programs?
Most university AI program deadlines fall between January and April 2027, so this winter is a great time to prepare. Here's when each one closes this cycle.
Stanford AIMI internship and bootcamp: The financial aid deadline is Jan 15, 2027, and the standard deadline is Feb 1, 2027.
Stanford AI4ALL: Feb 1, 2027 (AI4ALL admissions).
Carnegie Mellon AI Scholars: Feb 1, 2027, with decisions on Apr 1.
MIT BWSI: The online course opens on Feb 1, 2027, and recommendations are due on Apr 2.
MIT Jameel Clinic bootcamp: Regular applications close on Feb 22, 2027.
If you're too young for these programs, you're not behind. Younger students can start building AI skills now with a small project of their own. Our tips on how to find AI research opportunities in high school can help you pick a first step.
How to apply to a university AI research program
A strong application shows what you've already tried in AI and why the program's topic interests you. Here are 4 tips.
Apply to more than one program. Four of these programs close on the same day in early February, so you can plan your essays together.
Show your coding. The AIMI internship asks for strong programming skills. Describe one project you built, what it did, and what you'd change.
Ask for recommendations early. Carnegie Mellon asks for two letters. Give teachers three or four weeks, and remind them what you did in class.
Finish the BWSI online course. BWSI says students who don't finish enough of it are unlikely to get in. Set aside time each week in February and March.
If the AIMI internship wants more coding than you've done, you can build that skill this winter. In Veritas AI's online AI Scholars program (separate from Carnegie Mellon's), you can learn Python from your very first class and then use it on a real AI project. You'll team up with a few classmates on it.
FAQs
Can you do AI research in high school without a university program?
Yes. Many students do AI research with a mentor or a teacher, and some write it up for a journal. Our guide on how to write an AI research paper in high school walks through each step. Comfortable with Python and have a question to test? The Veritas AI Fellowship could fit. You'll get 15 one-on-one sessions with a mentor to shape your own project. You'll also get 4 sessions with a publication specialist, who helps you prepare for a journal, a competition, or a showcase.
What kind of AI research can high schoolers do?
You can work on real problems in areas like computer vision, medical AI, robotics, and language. For example, students in MIT BWSI's Medlytics course use machine learning to spot cancer in mammograms and to sort sleep stages from body signals.
Are there free AI research programs for high school students?
Yes. Carnegie Mellon's AI Scholars is free for everyone it admits. MIT's BWSI is free for families with income and assets under $200,000. The other programs here offer need-based aid.
Are there AI research programs for 9th graders?
Yes. Stanford AI4ALL is only for 9th graders. The two Stanford AIMI programs take students entering 9th to 12th grade in the fall. That means current 8th graders can apply too.
Can international students apply to these AI programs?
Some can. The Stanford AIMI programs welcome students abroad who can join on Pacific time. Stanford AI4ALL takes international 9th graders too, and it offers them aid. Carnegie Mellon's AI Scholars and MIT's BWSI are only for US students, and the Jameel Clinic doesn't sponsor visas.
Sources
Each program's own page has the latest details.
Keep Building Your AI Research With Veritas AI
A summer program can spark an AI question, and you don't have to wait for June to start answering it. More than 3,200 students in over 160 countries have learned AI at Veritas AI. They've worked with a team of 200+ mentors, including researchers at Anthropic, Meta, and NVIDIA. With a Veritas AI mentor, you can turn an AI question of your own into a finished research project. So far, 79% of the papers students have submitted have been published, according to Veritas AI internal data. You can compare Veritas AI's programs to see which one fits your coding level.
About the author: Tyler Moulton is the General Manager of the Lumiere Research Scholar Program and a former Harvard and Cambridge researcher in AI and astronomy.

