Berkman Klein Center Fellowships

The Berkman Klein Center for Internet & Society at Harvard University is now accepting fellowship applications!

Applications are now open for scholars and practitioners who wish to hold a fellowship with the Berkman Klein Center (BKC). We seek candidates who will propose and lead independent research initiatives aligned with BKC’s interdisciplinary AI research agenda.

Fellows appointed through this call will bring enthusiasm for working in interdisciplinary and intersectoral environments; fluency in communicating and translating between technical and non-technical stakeholders and audiences; excitement about working with and mentoring students; and a shared commitment to BKC's public interest mission and to open-source, accessible AI research.

We strongly encourage fellows to be in residence in Cambridge, MA, although non-resident fellowships will be considered on a case-by-case basis.

We welcome applications for two distinct appointment periods: 

• January-August 2026
• 2026-2027 Academic Year (September 2026 - August 2027)

More information about our call for applications is detailed below.

Applications will be accepted until Friday, December 5, 2025 at 11:59 p.m. ET. 

Please see additional application instruction information.

About the BKC Fellowship Program

Since its founding in 1996, BKC has taken a unique approach to developing and delivering innovation in modes beyond the confines of a traditional university. This is due in large part to the unusual model the center has adopted and honed for fellowships. While traditional university programs emphasize and rely on academic credentials to identify fellowships, the BKC Fellowship program considers and values a wide range of experiences, credentials, and potential contributions, with an emphasis on multisectoral candidates and interdisciplinary approaches to research and real-world impact.

Research Priorities

For this fellowship cycle, the Berkman Klein Center will prioritize research to inform and shape the design, development, and governance of AI systems pertaining to the following sets of issues:

Agentic AI

The deployment of agentic AI represents a change in kind: from passive chatbots and assistants to active participants in social, economic, and political processes. Unlike chatbots that interact with a human user, AI agents pursue objectives across time, modify their environments, and increasingly interact with other agents – without human mediation. This transition is occurring rapidly and haphazardly, and three critical gaps define this moment. First, we lack mechanisms for fine-grained measurement and control of AI agent behavior. Second, heterogeneous, ad-hoc, multi-agent systems will likely produce emergent behaviors we cannot understand, predict, or govern. Third, deployment is outpacing institutional adaptation and governance. We have no frameworks for agent accountability or liability, no models for AI economic participation, and no consensus on protections for humans in human-agent interactions. These gaps compound: without measurement, it is much harder to regulate well; without understanding multi-agent dynamics, we cannot prevent harms to people and cascading systemic failures; without institutional and governance frameworks, deployment patterns will entrench before we understand their consequences.

Language Model Interpretability

Language models have remarkable capabilities while remaining fundamentally opaque. We can observe what they do, but we do not generally know how or why they do it. This represents more than a scientific curiosity; it undermines meaningful oversight and safe deployment in systems increasingly embedded in high-stakes decision-making. We view interpretability as a (socio-)technical and institutional challenge, and seek to develop new methods to probe model internals while simultaneously building frameworks for interpretability standards and audit requirements that are actionable for researchers, policymakers, and users.

Benchmarking AI Systems Beyond Measures of Intelligence

AI systems continue to saturate benchmark after benchmark, but we are left with an unresolved question: are we measuring and controlling what actually matters to us? Measures of “intelligence” are too narrow to answer most of the questions we care about. To move forward, we must broaden our focus to include the non-intelligence aspects of computational systems, such as agency, identity, loyalty, metacognition, theory of mind, social cognition, situatedness, awareness, and even subjective experience. By developing benchmarks and interventions directed at these non-intelligence dimensions of computational systems, we aim to provide technologists, policymakers, and the general public with the empirical evidence needed to ground their positions and the control mechanisms to effectively and safely govern increasingly capable AI systems.

AI & the Human Experience

We explore how our increased reliance on AI is already changing and could transform core dimensions of being human. We are seeking to understand how AI will impact human relationships and connections, cognitive capacity and creativity, spirituality and faith, and social-emotional development. Our work aims to evaluate the extent of these impacts and to develop concrete legal, policy, and other interventions to address them. This work centers on the experience of being a human being—agency, dignity, community, meaning, and purpose—and develops actionable mechanisms to steer AI in ways that affirm our humanity rather than erode it.

Bridging the AI Triad

We are bringing together three foundational but typically siloed communities in AI: accelerationists, who often view AI as a revolutionary force for human progress; safetyists, who emphasize its potentially catastrophic or existential risks; and skeptics, who see AI as an incremental, over-hyped technology that yet carries dangerous near-term harms. By opening up dialogue among these groups, we seek to foster understanding, encourage collaboration, and lay the groundwork for more thoughtful policy and technical development around AI.

Our Collaborative Approach - Opportunities and Expectations

The specific expectations for participants in the fellows program will be unique to each fellow, with two broad expectations outlined below.

Producing a Project that Contributes to Public Scholarship:

Fellows will produce at least one significant public output that impacts and/or informs the scholarly, public, and/or policy debates in the arenas in which they work and BKC’s research agenda. These outputs could take many forms, including:

• Technical or design prototypes
• Novel machine learning techniques and algorithms
• Open-source research tools and datasets that advance the broader AI research community
• Public writing or audio/visual content, such as long-form pieces, op-eds, blog posts, policy briefs, podcasts, TED-style talks, or video shorts
• Academic writing, such as research papers, reports, or white papers
• Workshops or other convenings organized and led by the fellow with a public output of some kind 

Engaging with BKC Community Programming

Fellows will engage with faculty, staff, students, and other members of the BKC and Harvard University communities to learn with and from others and strengthen their own work. BKC’s generous community, built with intention and care over many years, is one of the Center’s great assets. Fellows activate this far-reaching network through events, workshops, listserv dialogues, reading groups, joint projects, and more.

Time and Location Commitments

Fellowships Between January 2026 - August 2026

These fellowships will last a period of up to eight months between January 2026 and August 2026. Specific dates and commitments will be discussed and determined between the fellow and the Berkman Klein Center.

Fellowships in the 2026-2027 Academic Year

These fellowships will run from September 1, 2026 to August 31, 2027

Applicants may opt to be considered for either or both of the time periods.

BKC strongly encourages fellows to be in residence in Cambridge, MA for a majority of their appointment, although non-resident fellowships will be considered on a case-by-case basis. During the time spent in residence, fellows will be invited to work from the Berkman Klein Center’s offices on the Harvard Law School campus. Fellows are expected to be free of the majority of their regular commitments so that they may fully devote themselves to their fellowship. We recognize that fellows who bring their own funding might have specific commitments due to their external arrangements.

Who Should Apply?

The Berkman Klein Center is a space for both established and rising scholars and practitioners from across disciplines and backgrounds. We seek candidates who have a demonstrated record of contributing to public and scholarly conversations and taking action, whether in the realm of policy, technology development, academia, and/or civil society. BKC seeks candidates eager to deploy their work in service of understanding and advancing the public interest.

Disciplines

Our fellows represent the full range of disciplinary backgrounds, from technology and industry, to law and policy, to the applied and social sciences, to the arts and humanities. Collectively, we foster research, dialogue, and building that bring many perspectives and methods together to broaden understanding and solve real-world problems. While we welcome experimental and non-traditional research, candidates should have experience in carrying out the form of work they propose to undertake during their fellowship. We particularly welcome candidates with interdisciplinary backgrounds who blend technical and non-technical expertise.

• For candidates primarily interested in scientific research who wish to propose and lead independent AI research aligned with our research priorities, we strongly encourage applicants to apply with:
  • A Ph.D. in Computer Science or related technical field, or equivalent practical experience
  • Demonstrated expertise in Python and modern AI/ML frameworks (e.g., PyTorch, JAX)
  • Primary author publications in peer-reviewed Computer Science conferences or equivalent technical contributions
  • Ability to communicate with both technical and non-technical audiences
• For candidates primarily interested in research engineering who wish to propose and lead the development of open-source AI research infrastructure, we strongly encourage applicants to apply with at least three of:
  • Advanced degree in Computer Science or related technical field, or equivalent practical experience
  • Demonstrated expertise in Python and modern AI/ML frameworks (e.g., PyTorch, JAX)
  • Familiarity with modern agent frameworks (e.g., DSPy) and communication protocols (e.g., MCP, A2A)
  • Experience with HPC workload management systems (e.g., Slurm) and modern orchestration systems (e.g., Kubernetes, Ray, Airflow) on local machines and in cloud providers
  • Hands-on experience with open-weight models and the infrastructure required to train, evaluate, and serve them
  • Track record of building reproducible research infrastructure and experiment tracking systems (e.g., MLflow) 

Academics

We welcome applications from faculty for whom serving as a professor is their full-time commitment (including assistant, associate, and full professors or equivalent roles in countries outside of the U.S.) and post-doctoral scholars who have recently received a doctoral degree or other terminal degree by the start of their appointment.

Practitioners

We welcome applications from practitioners who have built their careers and research outside of academia, in areas such as industry, government, and/or civil society.

International Applicants

We work with the Harvard International Office (HIO) to sponsor visa paperwork for our eligible international fellows. An outline of the visa application process and requirements may be found on the HIO website at: http://hio.harvard.edu/scholar-visa-process.

Support

Stipend

The Berkman Klein Center has a limited pool of funding to support fellows, and funded fellowships, whether partial or full, are extremely competitive. Candidates may apply to be considered for fellowship funding from BKC, or to be considered for a fellowship supported by external funding.

• Fellowship funding: Candidates taking unpaid leave from their home institutions or who do not have any other outside funding may apply for BKC funding. A fully funded fellow appointed through the open call for applications is eligible to receive a stipend of up to $6,250 per month, up to $75,000 for a 12-month period. Specific stipend arrangements will be determined on a case-by-case basis with selected candidates.
• External funding: Candidates on paid sabbatical from their home institution or who are otherwise supported by external funding, who do not require a stipend from the Berkman Klein Center to support their fellowship.

Important Notes:

• If one is based in the United States but is not a United States citizen or Lawful Permanent Resident (“green card” holder), one’s immigration status must allow for the receipt of a fellow's stipend.
• Fellows may be responsible for tax reporting on their stipends. Please review additional information about stipends issued through Harvard University.

Access to University Resources

• Space: For their time spent in Cambridge, fellows will be provided with shared office/work space. We endeavor to provide comfortable and productive spaces for coworking and flexible use by the community.
• Library Access: All fellows will be provided with access to Harvard’s extensive libraries and research facilities.
• Campus Resources: Fellows are welcome and encouraged to connect with Harvard University’s research centers, initiatives, resource groups, associations, organizations, and specialized offices.
• Courses: Fellows may seek opportunities to audit classes across Harvard University. However, they must ask for direct permission from the professor of the desired class.
• Teaching at Harvard: Fellows may be able to teach at one of several Harvard schools. This would be determined on a case-by-case basis, arranged directly by the Fellow in collaboration with the respective schools’ administrations. BKC cannot promise any teaching engagement during the program.
• Health Insurance: Fellows should review Harvard University Health policy to determine whether they are eligible to purchase health insurance through the university.

Community Principles, Policies, and Resources

The Berkman Klein Center community, and how we interact with one another, is governed by norms and policies developed and maintained by Harvard University and Harvard Law School. The Center maintains a page to highlight community principles, policies, and resources, as well as other applicable policies and resources for accessing additional University support.

Notice of Nondiscrimination

Harvard University and Harvard Law School do not discriminate against any person on the basis of age, race, color, national origin, sex (including gender identity and gender expression, as well as pregnancy), genetic information, ancestry, religion, caste, creed, veteran status, disability, military service, sexual orientation or political beliefs in admission to, access to, treatment in, or employment in its programs and activities.

Application

Applications will be accepted until Friday, December 5, 2025 at 11:59 p.m. ET.

In addition to a short personal and work-related questionnaire, applicants will be required to upload the following documents. Please consider this information carefully and ensure your attachments meet these requirements: 

• CV
• 1-2 page cover letter: Please briefly tell us about your background, motivations, and goals. Why is the Berkman Klein Center the right place for you to do this work? What skills, expertise, connections, and insights will you contribute to the Center’s activities and community? How will the opportunity to engage colleagues from different backgrounds stimulate and accelerate your work? If applicable, kindly alert us to any relevant deadlines at your home institution that might affect your ability to accept a fellowship appointment.
• 2-3 page project proposal: What is the research you propose to conduct during a fellowship year? Please describe the problems you are trying to solve, the methods that inform your research, and the intended audiences for your outputs. As you are able, please describe how it aligns with one or more of the Center’s research priorities.
• A PDF of 1-3 work samples: Ideally, these should connect to the project proposal in some way or help to demonstrate the feasibility of the project proposal. Please submit these samples as one combined PDF. Do not include more than three samples; we will only review the first three samples.
• The name and contact information for two professional references:  If considered as a fellowship finalist, we may contact references to receive letters of recommendation or to conduct reference calls.

Please note that all uploads need to be PDFs. Individual files must not exceed 5 MBs.