When does class start/end?
Classes begin promptly at 9:00 am, and typically end at 5:00 pm.
Understanding Generative AI Risks: Security, Ethics, and Social Implications
As Generative AI (GenAI) becomes more widespread and impacts critical processes, understanding and managing its risks is crucial for successful operations. This risks of Generative AI training...
Read MoreAs Generative AI (GenAI) becomes more widespread and impacts critical processes, understanding and managing its risks is crucial for successful operations. This risks of Generative AI training teaches attendees how to spot the dangers, risks, and vulnerabilities of GenAI. Participants learn to methodically examine GenAI systems for potential hazards and identify mitigations for those risks. In addition, attendees gain industry-leading resources to stay updated in this rapidly evolving area.
Skills Gained
- Grasp the fundamental ethical principles of Responsible AI, such as fairness, accountability, and transparency, and how they affect real-world AI scenarios
- Discover how AI risks emerge from violations of Responsible AI principles
- Understand industry-standard categorizations and mitigations of AI risks
- Differentiate risks of Generative AI and the cybersecurity risks posed by it
- Identify potential vulnerabilities to AI systems and their defenses
Prerequisites
This course is designed for personnel responsible for identifying, assessing, and managing the risks of Generative AI in their organization. It assumes they understand how Generative AI functions at a workflow level, including core steps in the training and prediction process.
Course Details
Materials
All Generative AI Risks training students receive comprehensive courseware.
Software Needed on Each Student PC
Students should have Zoom installed as the conference platform.
AI Ethics and Responsibility
- What is an AI system?
- The AI System Lifecycle
- Common AI Actors
- Principles of AI Ethics
- Safe
- Secure & Resilient
- Explainable & Interpretable
- Privacy-Enhanced
- Fair
- Accountable & Transparent
- Valid & Reliable
- Cybersecurity Triad & The Fallout of Failure
GenAI Risks & Mitigations
- Nefarious Information (e.g., CBRN)
- Hallucinations
- Dangerous or Violent Recommendations
- Data Privacy
- Environmental Impacts
- Human-AI Configuration (e.g., workforce impact)
- Information Integrity (e.g., misinformation)
- Information Security
- Intellectual Property
- Obscene, Degrading, and Abusive Conduct
- Toxicity
- Bias
- Homogenization
- Supply Chain Integration
GenAI Cybersecurity – Top 10 Vulnerabilities & Defenses
- OWASP LLM Top 10
- Prompt Injection
- Insecure Output Handling
- Training Data Poisoning
- Denial of Service
- Supply Chain Vulnerabilities
- Sensitive Information Disclosure
- Excessive Agency
- Overreliance
- Model Theft
GenAI Cybersecurity – Tactics, Techniques, and Mitigations
- MITRE ATLAS
- Reconnaissance
- Resource Development
- Gaining Access
- Execution
- Persistence
- Privilege Escalation
- Defense Evasion
- Credential Access
- Discovery
- Collection
- ML Attack Staging
- Exfiltration
- Impact
Conclusion
- Frontier Threats
- Additional Resources
- Responsibility Matters