THE AI Security Fund (AISF) aims to accelerate and expand the field of AI security research to promote the responsible development and deployment of frontier models, minimize risks, and enable independent and standardized assessments of capabilities and security.
The AISF is inviting applications for research funding and will award grants to accelerate research efforts identifying potential security threats arising from the development and use of cutting-edge AI models. As AI is increasingly applied across fields, the fund aims to develop strategies to avoid negative outcomes by funding research to identify and assess risks and improve the safe deployment of AI for the benefit of society.
Through the fund’s cybersecurity tender, the AISF will support technical research that assesses potential risks and develops security measures for AI systems operating in cybersecurity contexts. The funding aims to promote the responsible development of frontier AI models while establishing robust assessment frameworks for security-related capabilities and safety measures. This call for proposals will support technical research into advanced AI systems to reduce cybersecurity risks.
Although AI systems demonstrate their value in defensive cybersecurity applications, their potential to automate or enhance offensive capabilities requires careful evaluation. Key concerns include the ability of AI systems to identify vulnerabilities, generate exploit code, and adapt known attack patterns to new contexts.
The research objectives of the RFP include benchmarks and hardening studies to evaluate cutting-edge AI models: ability to identify and exploit new security vulnerabilities, effectiveness in automating complex attack chains and ability to adapt and modify existing operating code. Upgrade studies, interdisciplinary tests and forecasting studies will be supported
The AISF plans to fund research projects conducted by academic laboratories, nonprofit organizations, independent researchers, and mission-driven for-profit entities on cybersecurity-related topics, in a range of 350,000 $ to $600,000.
Proposal types and eligible applicants include technical research projects focused on evaluating and improving AI security in cybersecurity applications; projects should focus on frontier AI models and their deployed versions.
For full program guidelines and application instructions, visit the AI Safety Fund website.
