The Amazon AI PhD Fellowship Program is a $68 million initiative that funds more than 100 PhD students working in core artificial intelligence research: machine learning, computer vision, natural language processing and closely related areas. Fellows get two years of tuition, fees and stipend support, travel grants, AWS cloud-computing credits and a senior Amazon scientist as a mentor. That is a strong package. The catch is that it runs only through nine named US universities, and for most Pakistani readers that single detail decides everything.
This guide explains who can realistically use this fellowship, how the university-led nomination works, and what to prepare. For a compact summary, see the full scholarship details.
The nine universities that decide whether this is for you
Before reading anything else, check where you are enrolled. The program currently operates through these partner universities only:
- Carnegie Mellon University
- Johns Hopkins University
- MIT
- Stanford University
- UC Berkeley
- UCLA
- University of Illinois Urbana-Champaign
- UT Austin
- University of Washington
Amazon states explicitly that students of any citizenship or nationality are eligible. So a Pakistani PhD student already enrolled at one of these nine schools can apply on the same footing as anyone else there. A student who is not enrolled at one of them currently has no route into this program, whatever their nationality, grades or publication record.
That makes this fellowship different from most of the scholarships Pakistani students search for. You cannot apply to it from Lahore or Karachi as a fresh graduate. It is funding for people who are already inside a specific set of AI PhD programs in the United States.
| Item | Details |
|---|---|
| Provider | Amazon (Amazon Science) |
| Funding | Fully funded: two years of tuition, fees and stipend, plus travel grants and AWS credits |
| Level | PhD |
| Fields | Machine learning, computer vision, natural language processing, core AI research |
| Country | United States (nine partner universities only) |
| Deadline | Set by each partner university's internal nomination or application process |
| Funding period | 2025-2026 and 2026-2027 academic years |
| Official site | Amazon Science announcement |
What two years of Amazon funding actually includes
The direct funding covers tuition, fees and stipends for two years. On top of that, fellows receive travel grants for conferences and research-related travel. The source material describes the stipends as generous but does not publish a specific monthly or annual figure, so do not assume a number. Ask your department what the stipend would look like at your school.
The part that matters most for AI researchers is probably the compute. The program sets aside a substantial pool of AWS cloud-computing credits each academic year. If your work involves training large models or running big experiments, you already know how quickly a normal academic budget runs out. Credits at this scale let you run work that would otherwise be hard to afford.
Then there is the mentor. Each fellow is paired with a senior Amazon scientist working in a related research area, who meets with the fellow regularly to discuss the technical direction of the research and its real-world implications for Amazon's products and services. Purely financial fellowships do not give you this. For a student thinking about industry research later, it is a direct view of how a large technology company turns academic AI work into deployed systems, and it can be a bridge to future Amazon research internships or scientist roles.
The whole program is part of a $68 million commitment funding over 100 PhD students across two academic years.
Eligibility beyond the university list
Enrollment at a partner school is the first filter. There are four more conditions:
- You must have passed your home department's required PhD qualifying exams.
- You must have an assigned PhD thesis advisor.
- Your research must sit within core AI disciplines (machine learning, computer vision, NLP or related areas).
- Any citizenship or nationality is accepted.
The qualifying exam rule is the one first-year students trip over. If you have just started your PhD and have not cleared quals or settled on an advisor, you are not ready for a nomination yet. Plan for a later cycle instead, and confirm with your department whether the program will still be running when you qualify, since the stated funding covers the 2025-2026 and 2026-2027 academic years.
The field rule also deserves attention. If your work is in systems, theory or another area of computer science that only touches AI at the edges, you will need to show clearly how it connects to core AI research. Tangential projects are a weak fit.
How the departmental nomination works
There is no central Amazon portal for this fellowship. Amazon runs it through the partner universities, and each one sets its own nomination or internal application process and its own deadline. In practice that means the steps look like this:
- Confirm you are enrolled in a qualifying AI-related PhD program at one of the nine partner universities.
- Ask your department or graduate school how they handle the Amazon Fellowship nomination or application.
- Prepare your research statement, CV and advisor confirmation in the format your department asks for.
- Submit through your university's internal process before its deadline.
- If selected, you are paired with an Amazon scientist mentor and receive funding for the applicable academic year or years.
The materials are fairly light compared with a typical scholarship file:
- A departmental nomination or internal application, following your university's process
- A research statement or summary of your ongoing PhD research
- A CV covering your academic and research background
- Advisor confirmation that they supervise your thesis
Because the departmental process controls both the format and the timing, treat your graduate coordinator's instructions as the final word. They may ask for more than this list.
Timing: why there is no single deadline
The program was announced in Fall 2025. University-level nomination processes vary by institution, and funding covers the 2025-2026 and 2026-2027 academic years. That is the full timeline in the available information. There is no published date that applies to every school, so a deadline you see quoted for one university tells you nothing about another.
The practical advice is simple: contact your department early. Internal nominations often close well before anyone outside the department notices, and if you only start asking once a friend mentions it, you may already have missed your school's window. Check the official Amazon announcement for program-level updates and your department for the actual date.
What applicants get wrong
The most common misunderstanding is thinking this is a scholarship you can use to get into a US PhD. It is not. It funds students who are already enrolled at the partner schools. If you are still choosing where to apply for a PhD, you can reasonably factor this fellowship into your thinking when AI-focused funding matters to you, but do not choose a university on the strength of this program alone. The eligibility conditions (quals passed, advisor assigned) mean you would only become eligible after you are settled in, and funding is tied to specific academic years.
A second mistake is pitching a broad or loosely AI-related project. Amazon is funding core AI research, so your research statement should make the machine learning, computer vision or NLP angle obvious in the first paragraph. If your advisor can help you frame the work in those terms, ask for that help.
Third, students sometimes skip the question of competition. Ask your department how many students Amazon typically funds at your school per cycle. That number tells you how hard the internal round will be and whether your time is better spent polishing this application or another one.
Finally, make sure your status is in order before seeking a nomination. Passed qualifying exams and a confirmed advisor are requirements, not nice-to-haves. A department will not nominate someone who does not meet them.
Who should skip this one
If you are a Pakistani student hoping to start a PhD in the US and need funding to make that happen, this fellowship will not help you now. You are better served by funding that accepts applicants from outside the US. The all United States scholarships guide and the broader scholarships guide for Pakistani students are better starting points.
If you are already in an AI PhD at one of the nine schools, past quals and working with an advisor, this is one of the stronger funding options you can pursue, mainly because of the compute credits and the mentorship on top of the money.
Frequently asked questions
Can a Pakistani student apply from Pakistan?
No. Pakistani students are eligible, because Amazon accepts any citizenship or nationality, but only if they are already enrolled in a PhD program at one of the nine partner universities. There is currently no route in from outside those schools.
How much is the stipend?
The fellowship covers two years of tuition, fees and stipend, plus travel grants and AWS credits. A specific stipend amount is not published in the available information, so confirm the figure with your department.
Is there a central Amazon application portal?
No. Each partner university runs its own nomination or internal application process with its own deadline.
Can a first-year PhD student apply?
Only if they have already passed the department's PhD qualifying exams and have an assigned thesis advisor. Many first-year students will not meet both conditions yet.
Which research areas count?
Core AI disciplines, including machine learning, computer vision and natural language processing, along with related areas. Work that only touches AI tangentially is a weaker fit.
For program updates, read the official Amazon Science announcement, and keep the Amazon AI PhD Fellowship scholarship page handy for the key details.