A newly released open-access paper written by representatives from academia, industry, and the Software Engineering Institute (SEI) says that the general-purpose artificial intelligence (GPAI) ecosystem needs to learn lessons from the software security domain to keep AI systems secure.
The paper ‘In-House Evaluation Is Not Enough: Towards Robust Third-Party Flaw Disclosure for General-Purpose AI’ raises the alarm about GPAI safety and the need for coordinated vulnerability disclosure (CVD). The paper notes that the past year’s proliferation of GPAI systems, which it defines as ‘foundation model-based software systems, with a wide variety of uses’, has outpaced the “infrastructure, practices, and norms for reporting flaws.” The authors propose standardized GPAI flaw reporting, the adoption of safe-harbor disclosure programs by GPAI system providers, and improved disclosure infrastructure.
Abstract (verbatim)
The widespread deployment of general-purpose AI (GPAI) systems introduces significant new risks. Yet the infrastructure, practices, and norms for reporting flaws in GPAI systems remain seriously underdeveloped, lagging far behind more established fields like software security. Based on a collaboration between experts from the fields of software security, machine learning, law, social science, and policy, we identify key gaps in the evaluation and reporting of flaws in GPAI systems. We call for three interventions to advance system safety. First, we propose using standardized AI flaw reports and rules of engagement for researchers in order to ease the process of submitting, reproducing, and triaging flaws in GPAI systems. Second, we propose GPAI system providers adopt broadly-scoped flaw disclosure programs, borrowing from bug bounties, with legal safe harbors to protect researchers. Third, we advocate for the development of improved infrastructure to coordinate distribution of flaw reports across the many stakeholders who may be impacted. These interventions are increasingly urgent, as evidenced by the prevalence of jailbreaks and other flaws that can transfer across different providers’ GPAI systems. By promoting robust reporting and coordination in the AI ecosystem, these proposals could significantly improve the safety, security, and accountability of GPAI systems.
