forked from rohitg00/ai-engineering-from-scratch
-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy pathquiz.json
More file actions
78 lines (78 loc) · 2.73 KB
/
Copy pathquiz.json
File metadata and controls
78 lines (78 loc) · 2.73 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
{
"lesson": "26-compliance-frameworks",
"title": "Compliance — SOC 2, HIPAA, GDPR, PCI-DSS, EU AI Act, ISO 42001",
"questions": [
{
"stage": "pre",
"question": "When does EU AI Act enforcement for high-risk systems begin?",
"options": [
"February 2, 2025",
"August 2, 2026",
"Already fully enforced in 2024",
"January 1, 2030"
],
"correct": 1,
"explanation": ""
},
{
"stage": "check",
"question": "Which two-tier fine ceiling does the EU AI Act define?",
"options": [
"Up to €1M flat for any violation",
"Up to €100K for any violation",
"Up to €15M or 3% global annual turnover for high-risk-system obligations (Art. 99(4)); up to €35M or 7% for prohibited AI practices (Art. 99(3))",
"No financial penalties, only takedown orders"
],
"correct": 2,
"explanation": ""
},
{
"stage": "check",
"question": "Why is post-processing PII cleanup not a defensible GDPR posture?",
"options": [
"It is too slow at scale",
"GDPR forbids redaction entirely",
"Post-processing is identical to real-time",
"The model already saw the data, so real-time inference-layer redaction (before the LLM call) is the defensible 2026 standard"
],
"correct": 3,
"explanation": ""
},
{
"stage": "check",
"question": "What is the practical difference between SOC 2 Type I and Type II?",
"options": [
"Type I requires HIPAA BAA",
"Type I is more rigorous than Type II",
"Type I attests controls designed and documented; Type II attests controls operating effectively over 6-12 months",
"Type II is for startups only"
],
"correct": 2,
"explanation": ""
},
{
"stage": "post",
"question": "What does cross-framework control mapping aim to deliver?",
"options": [
"One control policy that satisfies multiple framework requirements (e.g. access logging maps to ISO 27001 A.5.15-5.18, GDPR Art. 32, HIPAA §164.312(a))",
"Eliminating audits",
"Replacing all frameworks with ISO 42001",
"More distinct controls per framework"
],
"correct": 0,
"explanation": ""
},
{
"stage": "post",
"question": "What does the lesson recommend for HIPAA + LLM workloads?",
"options": [
"HIPAA does not apply to LLMs",
"Never send PHI to an external AI service without a signed BAA; all three hyperscalers and major LLM API providers offer BAAs",
"Use only on-prem models, never managed",
"Ship PHI to any provider; BAA is optional"
],
"correct": 1,
"explanation": ""
}
]
}