Nine layered interventions can reduce the risk of AI-enabled biological attacks, according to a new RAND Corporation report. The strategy emphasizes a defense-in-depth approach to prevent misuse of biology-focused models. It targets specific vulnerabilities in data access and synthesis. Practitioners must now integrate these safeguards into model deployment to mitigate high-consequence security threats.