APPLICATION | FELLOWSHIP IN AI ETHICS – Forms and Applications

AI ethics is having a moment. Not the fun, cocktail-party kind (unless your cocktail parties involve procurement rules and training data). The serious kind: laws are tightening, public sector algorithms are under scrutiny, and every lab building something “intelligent” is quietly realizing they also need to explain what “good” looks like.

That’s why this Harvard Center for Hellenic Studies (CHS) Fellowship in AI Ethics is so intriguing. It’s not a giant research grant with a budget that comes with its own gravitational pull. It’s a one-year postdoctoral appointment (July–June) designed to buy you something rarer than money in academia: protected thinking time, plus Harvard online resource access for a full year, plus a 10-day educational trip to the United States to present and connect.

The stipend is $12,000, which is modest by “rent in 2026” standards. But the fellowship is strategically powerful if your project needs: (1) credibility, (2) access to paywalled journals and databases, (3) mentorship and a scholarly home, and (4) a tight, publishable output you can complete within 12 months.

And the Greece angle is not decoration. CHS is based in Greece and wants fellows who can connect AI ethics to humanistic and social questions in ways that resonate locally and regionally—policy, civil society, culture, institutions, history, language, all of it. If your work can naturally speak to Greek or Mediterranean contexts (or you can make a convincing case that it should), you’ll be playing the right sport on the right field.

At a Glance: Harvard CHS AI Ethics Postdoctoral Fellowship

What This Fellowship Actually Gives You (Beyond the Stipend)

Let’s be blunt: $12,000 won’t bankroll a lab. It won’t pay for a full year in most cities, and it won’t cover a grand fieldwork tour unless your fieldwork is primarily “reading and thinking intensely in the sunlight.”

But that’s not what this fellowship is selling. It’s selling concentration.

First, the stipend is meant to support living and research costs while you produce a discrete scholarly outcome. The key word is discrete. CHS is not asking for your life’s work. They’re asking for something you can finish and put into the world: a journal article, a policy brief with real bite, a book chapter, a normative framework you can defend without sweating through your shirt.

Second, the Harvard online resource access is quietly huge. If your home institution has thin library subscriptions—or you’re between affiliations, as postdocs so often are—this is like being handed a master key. Journal access, databases, digital archives: suddenly your literature review stops being an exercise in “I found the abstract and a pirated PDF from 2018.”

Third, there’s mentorship from an academic committee. Done well, this is the difference between “I worked on a topic for a year” and “I published something that moved my career forward.” You get feedback, positioning, and (often) introductions—the professional version of being waved through the gate instead of stuck outside explaining yourself.

Finally, the 10-day educational visit to the US is short, but it’s not a token. Ten days is enough to present work, meet scholars, and convert your project from “private obsession” to “public conversation.” If you’ve ever tried to build an academic network purely through email, you know why face time still matters.

Who Should Apply (And Who Should Probably Not)

This fellowship is for postdoctoral researchers who are ready to produce a meaningful AI ethics contribution in a year. It is emphatically not a “someday I might write about AI ethics” opportunity. The single-award format also means it’s competitive—one slot, many smart people, only one person gets the keys.

You’re a strong fit if you’ve already got the bones of a project and need the conditions to finish it. That might describe a philosopher clarifying value pluralism in AI design; a lawyer mapping how procurement and accountability collide in public-sector algorithms; an STS scholar studying how automated systems reshape institutional trust; or a computer scientist with an ethics spine who can explain social implications without hiding behind math.

CHS is interdisciplinary, which is refreshing—and also a trap for applicants who think interdisciplinarity means vagueness. Interdisciplinary here means you can speak to more than one audience without turning your proposal into a buffet of buzzwords.

Projects with a Greece or regional connection tend to make intuitive sense for this program. Examples that fit the fellowship’s personality include: examining algorithmic decision-making in Greek public services; comparing AI governance approaches among EU Mediterranean states; analyzing automated systems and migration pressures across the Aegean; or connecting ancient and modern ethical vocabularies to present-day AI debates in a way that isn’t gimmicky.

On the flip side, if your proposal is purely technical—say, optimizing a neural architecture with no ethical argument, legal analysis, or social consequence spelled out—this is likely not your best home. You can absolutely bring technical work, but the ethical stakes must be the point, not an afterthought tacked on at the end like parsley.

Eligibility and What Reviewers Are Really Checking

The formal eligibility is straightforward: you need a PhD (or equivalent) and must have defended by December 1, 2025. You must be fluent in English. Some modern Greek helps because the program wants genuine engagement with Greek partners and public conversations, but it isn’t automatically disqualifying if you don’t have it—especially if your project doesn’t require Greek-language materials.

The location note matters: being based in Greece is preferred, but not strictly required. If you’re not planning to be in Greece, you’ll need to explain how you’ll still participate meaningfully. “I will occasionally email someone” is not a plan. A plan looks like scheduled virtual seminars, periodic visits if feasible, a named partner institution, or a clear public-engagement route that aligns with CHS goals.

Selection-wise, committees like this are usually judging three things at once:

1. Is the idea sharp and important?
2. Can this applicant actually finish it in one year?
3. Does CHS in Greece make this project better (or even possible)?

Your job is to answer all three without wandering past 1,000 words.

What a Strong Project Looks Like in This Program

A strong CHS AI ethics project tends to have a spine, not just a topic. “AI and fairness” is a topic. “How should Greek public agencies justify and audit automated eligibility decisions under EU governance norms, and what counts as a meaningful explanation in Greek administrative practice?” is a project.

The best proposals also name a deliverable that sounds like it belongs on a CV line. Think: “Article submitted to [X journal] by May,” or “Policy brief plus working paper circulated to [specific audience].” It’s less glamorous than saying “I will rethink ethics,” but far more fundable.

And because this is the Center for Hellenic Studies, you can sometimes win points by demonstrating cultural and institutional literacy. Not by name-dropping Plato. By showing you understand how ethics arguments land in real contexts: courts, ministries, universities, NGOs, the public square.

Insider Tips for a Winning Application (The Stuff People Learn the Hard Way)

1) Treat the first 200 words like a movie trailer

Reviewers are busy, and a 1,000-word cap encourages skimming. Start with your research question, your core claim, and your intended output. If the opening paragraph could be swapped with any other AI ethics fellowship, rewrite it.

A simple formula works: problem → why now → what you will produce → why CHS.

2) Make CHS necessary, not merely nice

A lot of people can write “I would benefit from Harvard resources.” Fewer can say which resources and why. Name the types of materials you’ll rely on (journals, archives, databases), describe how you’ll use the US visit (presentation, workshops, meetings), and explain the Greece connection like you mean it.

If Greece is central—policy context, collaborators, field site—say so. If it’s not central, be honest and show how you’ll still contribute to CHS goals.

3) Pick one deliverable and defend it

This fellowship rewards focus. A single strong paper beats a fantasy trilogy. Choose one main output and, if you must, one secondary output (for example, an academic article plus a shorter public-facing brief). Then explain what “done” looks like.

Even better: give the deliverable a working title. It signals seriousness and makes the project feel real.

4) Build a timeline that sounds like a human wrote it

Month-by-month plans don’t need to be elaborate, but they must feel plausible. Avoid “collect data, analyze, write” as if those are three buttons you press.

A credible year often looks like: early months for literature and data access; middle months for analysis and outlining; late months for drafting, revision, presentation, and submission.

5) Use your writing sample like a strategic weapon

The sample can be up to 10,000 words, which is enough to show your best self. Choose something polished that matches the kind of work you’re proposing. If your proposal is legal-ethical analysis, don’t submit a purely theoretical meditation that never touches institutions. If your proposal is interpretive cultural analysis, don’t submit a methods appendix that reads like an instruction manual.

If you have a shorter publication record, the sample is your chance to prove you can write something reviewers would actually want to read.

6) Pre-wire your referees, even though you are only submitting contact details

Because you’re not uploading letters at submission, applicants sometimes treat references casually. Don’t. Choose two people who can speak to (a) your ability to finish, and (b) the specific project’s feasibility. Then send them a one-page summary with your thesis, deliverables, and timeline so they’re prepared if CHS contacts them.

7) Write for an intelligent outsider

This is interdisciplinary by design. That means at least one reviewer may not live in your subfield. Define key terms briefly. Explain why your method fits your question. If you must use jargon, make it earn its rent.

A good test: give your draft to one smart colleague outside your field. If they can’t summarize your project after reading it once, your proposal is too cloudy.

Application Timeline: A Realistic Workback Plan to February 6, 2026

The deadline is February 6, 2026, and the smartest applicants treat that as a soft deadline. You want to be done early enough that a portal glitch doesn’t decide your academic fate.

By mid-December 2025, aim to have your project framing nailed down: research question, tentative argument, and the one deliverable you’ll produce. This is also when you should approach referees. People disappear during holidays, and nothing good happens when you’re chasing a professor on January 30.

In early January 2026, write a full draft of the proposal and choose your writing sample. Don’t overthink it: pick the strongest piece that aligns with your proposed work, then edit it ruthlessly so the first pages are excellent. Update your CV with a focus on signals of follow-through—publications, talks, completed projects.

By mid-January, you should be revising, not inventing. Ask two readers for feedback: one in your field (for accuracy) and one outside it (for clarity). This is where you’ll discover that your “obvious” framing is not, in fact, obvious.

By late January, convert everything to clean, searchable PDFs. Confirm the proposal stays within 1,000 words (bibliography excluded) and the sample stays within 10,000 words (including bibliography). Then do a dry run: open the PDFs on a different computer and make sure formatting didn’t break.

Finally, by February 1–3, submit. Give yourself a 72-hour buffer. It’s the cheapest insurance you’ll ever buy.

Required Materials (And How to Make Each One Count)

You’ll submit three PDFs and provide contact details for two referees. The list is short, which means each item carries more weight.

• Research proposal (PDF, ≤1,000 words, in English). Treat this as a miniature journal article pitch: question, stakes, approach, and what you will deliver. Use light headers to make it easy to scan. Include a short bibliography to show you know the conversation and aren’t reinventing the wheel.

• CV (PDF, in English). Keep it clean and relevant. For early-career scholars, two to four pages is common. Emphasize publications, works in progress with credible status, conference talks, relevant research experience, and any public engagement that fits AI ethics.

• Sample of work (PDF, ≤10,000 words, in English or Greek). Submit your best writing, not your longest. If you’re trimming from a chapter, don’t just cut to length—edit so the excerpt stands on its own with a clear arc.

• Two referees (contact details). Provide institutional emails and choose people who will respond promptly if contacted.

What Makes an Application Stand Out (The Unofficial Scorecard)

Standout applications usually combine ambition with restraint. The committee wants to fund an idea that matters, but they also want the quiet confidence of someone who finishes.

Expect your application to shine if it offers a crisp research question, a plausible one-year plan, and a deliverable with a clear audience. Add evidence—publications, prior work, methodological competence—and you’re making the reviewer’s job easy in the best way.

Also: CHS fit. If your proposal connects naturally to Greece—through policy, institutions, archives, language, collaborations, or public engagement—say so plainly. If it doesn’t, compensate with a specific plan for how your CHS affiliation will shape the work and how you’ll contribute to CHS intellectual life.

Common Mistakes to Avoid (And the Fixes)

One common mistake is writing a proposal that sounds like it could be sent to any fellowship on Earth. That’s a rejection waiting to happen. Fix it by making the CHS connection concrete: resources, networks, Greece-based engagement, and how the US trip fits your dissemination plan.

Another is overpromising. A year disappears fast, and $12,000 doesn’t buy infinite time. Fix it by narrowing to a single case study, a defined corpus, or a tight conceptual puzzle you can actually solve in 12 months.

People also sabotage themselves with a mismatched writing sample. If the sample doesn’t resemble the work you’re proposing, reviewers have to guess whether you can deliver. Fix it by choosing the piece that proves you can do this project, not merely that you are smart in general.

Finally, there’s clarity. Dense, discipline-specific writing can alienate interdisciplinary reviewers. Fix it with short definitions, clear transitions, and sentences that do one job at a time.

Frequently Asked Questions

Can I apply if I am not based in Greece?

Yes. Greece is preferred, not strictly required. But you should explain how you’ll stay engaged with CHS and any Greek partners in a meaningful, scheduled way. Vague promises won’t reassure reviewers.

Do I need to speak modern Greek?

English fluency is required. Modern Greek is helpful, especially for projects that involve local institutions or Greek-language sources. If you don’t have Greek, show why your project remains feasible and how you’ll handle any language-dependent parts.

What is the PhD cutoff date?

You must have defended your PhD by December 1, 2025. If your defense is after that date, assume you’re ineligible unless the official page states otherwise.

Is this fellowship only for philosophers?

No. CHS explicitly welcomes postdocs from philosophy, computer science, sociology, law, classical studies, and other fields—provided your work is genuinely about ethics and the human implications of AI.

How many awards are available?

One. Which means you should treat the application like a publication submission: tight argument, clean writing, no wasted space.

Can I apply to other CHS fellowships in the same cycle?

No—this program indicates you may not apply to other CHS fellowship programs during the same application cycle.

When will I hear back?

The listing does not specify a notification date. Plan for a review period of a few months after the February deadline and monitor CHS communications for updates.

Will the stipend cover all living and travel costs?

It’s intended to support living and research costs, but whether it’s “enough” depends on your situation and location. Think of it as partial support paired with affiliation, resources, and visibility.

How to Apply: Your Next Steps (Do This This Week)

Start by writing a tight 200–250 word project summary that includes your research question, your one-year deliverable, and why CHS in Greece is the right setting. This short summary becomes your compass—if a sentence in the proposal doesn’t serve it, cut the sentence.

Next, pick your two referees and email them that summary plus the deadline. Ask explicitly if they’re willing to be listed and potentially contacted. Then choose your writing sample and polish it like you’re submitting it to a journal, because in a way, you are.

After that, draft the 1,000-word proposal with simple internal headers (Aims, Significance, Approach, Timeline, Outputs). Keep your scope tight. Make your deliverable obvious. Make the CHS connection unavoidable.

Finally, convert everything to PDF early, double-check word counts, and submit with a few days to spare. The only thing worse than rejection is a rejected upload.

Apply Now and Full Details

Ready to apply? Visit the official application page here: https://forms-greece.chs.harvard.edu/fellowship-ai-ethics-application/