Cover of AI Platform and Security

Book 3 · Presentation

AI Platform and Security

Building and Securing AI Platforms in European Enterprises and Public Authorities

61 slidesPowerPoint · 3.8 MB

Download the deck (.pptx) Read the book online

All presentations / AI Platform and Security

All slides

Select a slide to see it larger; use the arrow keys to move through the deck.

  1. Slide 1: AI Platform and Security1
  2. Slide 2: Foundations2
  3. Slide 3: Introduction to AI Platform and Security3
  4. Slide 4: Introduction to AI Platform and Security4
  5. Slide 5: Usage Modes and the Platform: What Each Mode Demands5
  6. Slide 6: Usage Modes and the Platform: What Each Mode Demands6
  7. Slide 7: Regulatory Requirements as Engineering Constraints7
  8. Slide 8: Regulatory Requirements as Engineering Constraints8
  9. Slide 9: A Reference Architecture for AI in Large Organisations9
  10. Slide 10: A Reference Architecture for AI in Large Organisations10
  11. Slide 11: AI Decision Records, Model Cards and the Architecture Repository11
  12. Slide 12: AI Decision Records, Model Cards and the Architecture Repository12
  13. Slide 13: Data, Models and DevSecMLOps13
  14. Slide 14: Data Architecture for AI14
  15. Slide 15: Data Architecture for AI15
  16. Slide 16: Model Lifecycle Management: MLOps and LLMOps as Code16
  17. Slide 17: Model Lifecycle Management: MLOps and LLMOps as Code17
  18. Slide 18: Setting Up an Effective DevSecMLOps Loop18
  19. Slide 19: The DevSecMLOps loop19
  20. Slide 20: Risks and Vulnerabilities Across the DevSecMLOps Loop20
  21. Slide 21: Risks at each stage of the loop21
  22. Slide 22: Foundation Models, Retrieval and Grounding22
  23. Slide 23: Foundation Models, Retrieval and Grounding23
  24. Slide 24: Agentic Architectures and Orchestration24
  25. Slide 25: Agentic Architectures and Orchestration25
  26. Slide 26: Integrating AI into the Enterprise Landscape26
  27. Slide 27: Integrating AI into the Enterprise Landscape27
  28. Slide 28: Infrastructure28
  29. Slide 29: Infrastructure, Compute and European Digital Sovereignty29
  30. Slide 30: Infrastructure, Compute and European Digital Sovereignty30
  31. Slide 31: Cloud AI Services versus On-Premises AI: Different Challenges31
  32. Slide 32: Cloud AI Services versus On-Premises AI: Different Challenges32
  33. Slide 33: Hybrid Agentic Work: Self-Hosted and Cloud Models for Sensitive Information33
  34. Slide 34: Hybrid Agentic Work: Self-Hosted and Cloud Models for Sensitive Information34
  35. Slide 35: Security35
  36. Slide 36: Security for AI Systems36
  37. Slide 37: The AI threat surface37
  38. Slide 38: AI Security Engineering in Practice38
  39. Slide 39: Defence in depth for an AI system39
  40. Slide 40: When AI Is the Threat: AI-Enabled Attackers and the Collapsing Exploitation Window40
  41. Slide 41: The exploitation window then and now41
  42. Slide 42: When Your Own Agents Exceed Their Mandate: Containment That Does Not Depend on Intent42
  43. Slide 43: Containment that does not depend on intent43
  44. Slide 44: Delivery and Operations44
  45. Slide 45: Evaluation and Testing of AI Systems45
  46. Slide 46: Evaluation and Testing of AI Systems46
  47. Slide 47: Observability, Monitoring and Drift in Production47
  48. Slide 48: Observability, Monitoring and Drift in Production48
  49. Slide 49: Cost, Capacity and Sustainability of AI49
  50. Slide 50: Cost, Capacity and Sustainability of AI50
  51. Slide 51: From Pilot to Production: Scaling AI Across the Organisation51
  52. Slide 52: The industrialisation path52
  53. Slide 53: Conclusions53
  54. Slide 54: Nordvik Region and Hansa Mutual: Platform and Security Conclusions54
  55. Slide 55: Nordvik Region and Hansa Mutual: Platform and Security Conclusions55
  56. Slide 56: Conclusions: Building and Securing AI Platforms56
  57. Slide 57: Conclusions: Building and Securing AI Platforms57
  58. Slide 58: Appendices58
  59. Slide 59: Code Examples and Technical Implementations59
  60. Slide 60: Templates and Tools60
  61. Slide 61: AI Platform and Security61