By Ken Muthomi - Reporter
United States, 7 April 2026 - OpenAI, ChatGPT’s mother company, has announced the OpenAI Safety Fellowship, a research initiative that aims to bring together technology experts from around the world to help improve advanced artificial intelligence systems.
In a statement released on Monday, OpenAI said the fellowship will bring together researchers, engineers, and practitioners to conduct research aimed at strengthening the safety and alignment of both current and future advanced AI systems.
Key areas of focus during the fellowship include safety evaluation, ethics, robustness, scalable mitigations, privacy-preserving safety methods, agentic oversight, and high-severity misuse domains, among others.
The tech company says that the programme is set to run from 14 September 2026 until 5 February 2027.
“We are especially interested in work that is empirically grounded, technically strong, and relevant to the broader research community,” OpenAI said.
“Fellows will work closely with OpenAI mentors and engage with a cohort of peers. Workspace will be available in Berkeley alongside other fellows at Constellation(opens in a new window), though fellows may also work remotely,” it added.
The company says that the fellowship is open to applicants from diverse academic and professional backgrounds, including computer science, social science, cybersecurity, privacy, HCI, and related fields.
Key characteristics the company says it will prioritise during the selection process are research ability, technical judgement, and execution over formal credentials. Applicants will also be required to have a letter of reference.
Related articles
Successful applicants will benefit from monthly stipends, computing resources, API credits, and ongoing mentorship.
The company has, however, clarified that participants of the fellowship will not be granted access to OpenAI systems and will instead work primarily with external infrastructure such as open-source models and public APIs.
Interested applicants have been encouraged to make their applications before 3 May 2026, through the official form, with successful applicants expected to be notified by 25 July 2026.
“Fellows are expected to work full-time for the duration of the program and produce a substantial research output by the end, usually in the form of a paper submission, benchmark, or dataset,” OpenAI said.
“You don't need a PhD, prior ML experience, or published papers. Successful fellows from similar fellowships have had backgrounds in physics, mathematics, computer science, cybersecurity, and other quantitative fields,” it added.
Ad
Ad