Protecting AI Systems Against Data Poisoning

Data poisoning—where adversaries tamper with training data to corrupt model behavior—poses significant risks as AI adoption expands across critical sectors. Organizations without mechanisms in place to detect or prevent data poisoning are open to an avenue of attack that, once exploited, is difficult to remediate. Machine unlearning and model retraining are not always viable or effective solutions. In today's operational climate, where threat actors look to influence models and degrade the trust of users through incorrect behaviors, preventing data poisoning is more important than ever.  In this episode of the SEI Podcast Series, Julie Lawler and James Cunningham—AI security researchers at Carnegie Mellon University's Software Engineering Institute—discuss the growing threat of data poisoning in AI systems and highlight emerging mitigation strategies, including chain-of-custody controls.