Google, IBM & Meta Certificates — All 10,000+ Courses at 40% Off
One annual plan covers every course and certificate on Coursera. 40% off for a limited time.
Get Full Access
Learn about secure AI application development in this 23-minute tutorial from Google's Mihai Maruseac. Explore essential security practices for AI-powered applications, addressing common vulnerabilities that plague traditional software but manifest more rapidly and with greater risk in AI systems. Follow a practical development journey of an ML model from data collection through production deployment, with detailed focus on building secure software supply chains, establishing clear model provenance using SLSA standards, and implementing capability analysis to mitigate future risks. Discover techniques for tracing production inference issues back to potential dataset poisoning, and learn strategies for optimizing model retraining costs when ML frameworks are compromised by analyzing impact radius. Master the fundamentals of creating robust AI systems while maintaining security at every development stage.
