CAIS Philosophy Fellowship

A seven-month research fellowship aimed at clarifying risks from advanced AI systems

What is AI Safety?

As artificial intelligence becomes increasingly advanced, the importance of safety in the development and deployment of these systems becomes apparent. The nascent field of AI safety is naturally interdisciplinary, with technical researchers studying and improving the safety of current systems, policymakers developing regulations for these systems, and philosophers clarifying the conceptual space around AI safety. As an example of a philosophical contribution, Peter Railton’s lectures on Ethics and Artificial Intelligence provided insight into developing AI systems sensitive to moral considerations. Similarly, Nick Bostrom’s work on Superintelligence laid out some of the potential dangers involved in advanced AI. The CAIS philosophy fellowship seeks to explore the various philosophical considerations surrounding AI safety.

The Program

The CAIS Philosophy Fellowship will be a paid, in-person opportunity located in San Francisco, CA.
Over the course of the fellowship, researchers will attend seminars and guest lectures, work closely with advisors, and conduct independent research.
Fellows will receive a grant of $60,000 (or more, depending on competing offers), a housing stipend of $20,000, and be connected to other institutions as the fellowship concludes.
Previous experience with AI or AI safety is not required.


Example Research Avenues

Implementing Moral Decision-Making

How can we build systems that are more likely to behave ethically in the face of a rapidly changing world? Should we align advanced AI systems with human values, human intentions, or social processes? How can we incorporate multiple stakeholders and moral uncertainty into an AI’s decision-making in practice?

Risks, Dynamics, Strategies

What processes might shape the behavior of advanced AI systems? How could the development and proliferation of an AI system go awry? Could advanced AI systems pose an existential risk? What are strategies to address these risks?

Tensions in Designing AI Systems

What are the inherent tensions surrounding the development of AI and how will these tensions be resolved? What are the advantages and disadvantages of (de)centralized AI systems? How do competitive pressures undermine AI systems that are power averse or altruistic? For agents tasked with pursuing a broad set of goals, how can one avoid incentivizing agents to develop power-seeking tendencies?

Criticism of Existing Literature

Are there substantial flaws in the existing concepts, arguments, strategies regarding AI existential risk? Are there any risks that have been overlooked in the existing literature?


Guest Speakers


Institutions Interested in AI Safety Philosophers

After the program, we plan to connect fellows to the following organizations which have expressed interest in the skillset that our fellowship seeks to develop.



Philosophy Ph.D. student, or
Graduate of a philosophy Ph.D. program (professors are encouraged to apply)


Program Benefits

$60k grant for the 7-month period*
$20k housing stipend
Covered student fees
Full-time research opportunities at CAIS for top-performing fellows
Covered cost of travel

*We are potentially willing to match competing offers


Application process

Applicants will complete a written application including a writing sample, a personal statement, and any research or publications they may have. A select number of applicants will then be interviewed remotely.
Selected candidates will receive funding to travel to San Francisco for a visiting weekend.


October 7

Applications close

November 18

Visiting weekend

January 9 *

Program begins

August 4 *

Program ends

Dates are flexible for candidates, based on their availability

Have questions?

To contact us, fill out the contact form, or email us directly at