Coordinated Disclosure for AI: Beyond Security Vulnerabilities

Harm reporting in the field of Artificial Intelligence (AI) currently operates on an ad hoc basis, lacking a structured process for disclosing or addressing algorithmic flaws. In contrast, the Coordinated Vulnerability Disclosure (CVD) ethos and ecosystem play a pivotal role in software security and transparency. Within the U.S. context, there has been a protracted legal and policy struggle to establish a safe harbor from the Computer Fraud and Abuse Act, aiming to foster institutional support for security researchers acting in …

artificial artificial intelligence beyond context coordinated coordinated vulnerability disclosure cs.ai cs.cr cs.cy cvd disclosure ecosystem ethos flaws harm intelligence play process reporting role security software software security transparency vulnerabilities vulnerability vulnerability disclosure