LCSEE graduate series seminar on AI’s rapid evolution—from large language models to agentic systems—has magnified risks of exploitability, bias, and misinformation. This talk examines real-world examples of AI misuse and emerging strategies to safeguard ethical AI use.