Planning: DCL-EPSCOR: SaTC Frontier: Exploring the Synergy Between Generative AI and Cybersecurity

Generative AI systems, which can produce text, images, and other content, are increasingly commonly used. However, from a security and privacy perspective, these systems pose a number of risks. Generative AI systems can violate the property and privacy rights of the people whose data are used to build them. They can also be used to generate content that might trick people into making mistakes that affect their security, privacy, and beliefs. On the other hand, generative AI systems might be used to support security and privacy goals. They may be able to create synthetic data that helps detect and predict new attacks more rapidly, which in turn could be used to make safer software, policies, and security operations practices. This project's goal is to support the planning of a large-scale proposal to study the risks and benefits that Generative AI-based systems may pose to security and privacy, developing methods to increase their safety and usefulness for society. The project's goal is to develop a fully realized proposal in terms of the scientific questions, team and partnerships, and social impacts to be addressed. To do this the research team will start with project ideation and identification of key stakeholders in both the scientific and social impact realms. These efforts will be based on established methods for identifying and learning from those affected by projects, including interviews, workshops, and community-building meetups. The research team will also identify an external advisory board and evaluation team to guide the project as it goes forward. Through these efforts, the team will develop the knowledge, resources, and people required to create a high-quality, large-scale proposal. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.