Top AI Safety, Governance, Security, and Policy Fellowships for 2026

Explore Top AI Safety, Governance, Security, and Policy Fellowships for 2026

Artificial intelligence (AI) is quickly changing our world, affecting everything from how we live to how governments operate and how global security is managed. As AI technology gets more advanced, there’s a growing need for experts who can guide its development, ensure its safety, improve its security, and make sure it’s used responsibly. To meet this demand, several important fellowships and leadership programs are now open for applications in 2026. These opportunities offer funding, mentorship, chances to connect with experts, research support, and ways to advance your career in the field of AI policy, governance, security, and safety.

This article looks at some of the best opportunities available for professionals, researchers, students, engineers, policymakers, and future leaders who want to help shape the future of AI.

Talos Network Artificial Intelligence Policy Leaders Programme 2026

The Talos Network Artificial Intelligence Policy Leaders Programme is designed for experienced professionals aiming to influence AI governance in Europe. It seeks to build Europe’s ability to manage advanced AI systems safely and effectively. The program supports professionals with significant experience in policy, government, international organizations, academia, and industry.

Programme Tracks

The program offers two main tracks:

Placement Track

This track is for professionals with at least eight years of experience who wish to move into AI governance. Participants get a 12-month paid placement at an AI-focused organization. They receive a base salary of EUR 5,000 per month for placements in Brussels, with adjustments for other cities. The track also includes career transition support, mentorship, and networking opportunities.

Training Track

This track is for senior professionals with at least twelve years of experience who want to integrate AI governance and safety into their current roles. Participants receive advanced AI governance training, support for strategic positioning, access to high-level policy networks, and career coaching.

Programme Components

Key components of the program include a virtual European AI Policy Bootcamp from September 10 to October 22, 2026, with weekly sessions compatible with full-time work. There is also a Policymaking Summit in Brussels from October 3 to October 9, 2026, offering fully reimbursed participation, workshops, and networking with senior policymakers.

Additional Benefits

Fellows gain access to one-on-one mentoring, career coaching, conference funding support, coworking support, and connections within the Talos Network alumni community. They also receive strategic career planning assistance and build connections across government, think tanks, industry, and European institutions.

Black in Artificial Intelligence Safety and Ethics Fellowship Fall 2026

This 13-week remote fellowship supports Black researchers, practitioners, and future leaders in AI safety, security, and governance. It runs from September 8 to December 18, 2026.

Research Tracks

Participants can focus on one of three areas:

AI Alignment

This track covers topics like interpretability, oversight and control, agency research, and machine learning safety.

AI Security

This track focuses on cybersecurity, adversarial robustness, risk management, and threat detection within AI systems.

AI Governance

This track explores policy frameworks, standards, evaluations, and systemic risk analysis related to AI.

Programme Structure

The fellowship is divided into two phases. Phase one, from weeks 1-5, involves AI safety seminars, expert training, and skill development. Phase two, from weeks 6-13, centers on mentor-guided research and capstone project development. Potential outputs include research preprints, technical articles, and grant proposals.

Benefits

Participants receive expert mentorship, AI safety training, research experience, professional development, and access to AI safety networks. The program also offers career preparation opportunities.

Secure Program Synthesis Fellowship 2026

The Secure Program Synthesis Fellowship is a joint effort by Apart Research and Atlas Computing. It addresses the critical challenge of ensuring AI-generated software is safe and functions as intended. The fellowship runs from June to September 2026.

Focus Areas

Projects may involve:

Specification Elicitation

Developing methods to turn general requirements into precise, formal specifications.

Specification Validation

Creating tools to check if specifications are accurate and complete.

Specification-Driven Development

Building systems that create and test software based on formal specifications.

Adversarial Robustness

Investigating weaknesses and potential failures in AI-assisted software development.

Benefits

Fellows receive research mentorship, access to compute resources, API support, community engagement, and travel funding. They also get opportunities to participate in Demo Day and publish their research. The fellowship is fully remote, requiring 8-30 hours per week. The application deadline was June 14, 2026.

Stanford University Center for International Security and Cooperation Artificial Intelligence Fellowships 2026-2027

Stanford University’s Center for International Security and Cooperation (CISAC) offers prestigious fellowships focused on AI, international security, geopolitics, and strategic competition. These fellowships provide an academic year of research, faculty mentorship, and access to Stanford’s researchers. Participants engage in seminars and a collaborative research environment, with opportunities for professional development. The application deadline was June 8, 2026.

Open Fellowship Tracks

Two tracks are available:

• Artificial Intelligence in Strategic Stability and Military Competition Fellowship
• Geopolitics of Artificial Intelligence Fellowship

These fellowships are ideal for researchers, scholars, security experts, AI governance professionals, and international affairs specialists.

Constellation Visiting Fellowship Fall 2026

The Constellation Visiting Fellowship supports AI safety researchers with a three-to-six-month research residency in Berkeley, California. The program covers housing and travel, provides 24/7 office access, and offers meals on weekdays. Fellows benefit from networking with over 100 researchers and attending research talks and events. Visa support is available for international participants. Applicants must be working full-time on AI safety or related research. Applications closed on June 12, 2026.

Coefficient Giving Navigators Incubator – Artificial Intelligence Security Cohort

The Navigators Incubator is a leadership development program focused on addressing global catastrophic risks, with its first cohort concentrating on securing transformative artificial intelligence. Participants receive salary matching support, research funding, compute resources, conference budgets, and networking opportunities. The program duration ranges from six months to two years.

Focus Areas

Participants can work on AI model security, infrastructure protection, compute verification, AI integrity, AI control, and rogue deployment detection. The program offers locations in Washington, D.C., San Francisco, London, and Tel Aviv, with remote participation also available. Ideal candidates include cybersecurity professionals, national security experts, AI security researchers, and technical founders. Applications are currently open.

Center for Human-Compatible Artificial Intelligence Research Fellowship

The Center for Human-Compatible Artificial Intelligence (CHAI) at UC Berkeley offers a research fellowship for postdoctoral researchers focused on developing beneficial and human-compatible AI. Fellows work closely with leading AI researchers and collaborate with Berkeley AI Research. The fellowship is intended for those aiming to create AI systems aligned with human values.

Research Areas

Potential research topics include theoretical foundations of AI, probabilistic programming, long-term decision making, human-robot interaction, and multi-human value alignment. The fellowship provides funding for two years, with a possibility of renewal, and a competitive salary. Visa sponsorship and cost support are available.

Eligibility and Application

Applicants need a PhD in a related technical field and a strong research record. Preferred qualifications include a solid understanding of modern AI and statistical learning. Application requirements include a CV, a one-page statement of research interests, and contact details for two academic referees. Applications are accepted on a rolling basis.

SHARE THIS POST WITH OTHERS

Frequently Asked Questions