Data Poisoning & Integrity Attacks

This lesson introduces data poisoning and integrity attacks, which target AI systems during the training process rather than after deployment. Learners explore how attackers manipulate training data to influence model behavior, create hidden vulnerabilities, or reduce system accuracy. The lesson examines poisoning techniques, backdoor attacks, data integrity risks, and defensive controls used to protect AI training pipelines. Students will gain a practical understanding of why trusted data is essential for trustworthy AI.