Risks and Mitigations for a Safe and Responsible AI
summary
Concrete claims and concrete artifacts matter: a risk framework that literally says Map Measure Manage, checklists tied to the EU AI Act, and measurable mitigations such as model selection, fine‑tuning for safety, monitoring and human in the loop shutdowns. Expect examples of prompt injection, hallucinations and data leakage, tests for dataset, association, interaction and confirmation bias, plus AI red teaming results, numbers on false‑positive/false‑negative rates, and whether safety brakes actually stopped harm.
AI-written summary · about
chapters (14)
- 0:00Introduction
- 1:25AI professionals' role
- 3:13AI risk frameworks
- 5:39Risk framework stages
- 8:47Trustworthy AI traits
- 10:58Privacy and adversarial risks
- 12:21Transparency and accountability
- 13:51Fairness and bias mitigation
- 17:25Bias in AI examples
- 21:19EU AI Act compliance
- 22:24Generative AI risks
- 23:47Mitigating generative AI
- 26:38Responsible AI culture
- 27:24Q&A
related talks
52mThe Risky Business, of AI Illiteracy
53mAI: Law, Policy and Common Sense Suggestions on How to Stay Out of Trouble
37mFundamentals of AI safety and security
37mRoadmap to Responsible AI
36mResponsible AI: In Theory, Practice, and Policies
30mIntroduction to AI Safety. Let's be ready before it's too late!
12mHow Varied Perspectives Drive AI Excellence
44mWhat Security Teams Need to Know and Do