ABCDEFGHIJKLMNOPQRSTUVWXYZ
1
2
GuideNet AI Control Framework
3
4
Overview and Usage Guide
5
6
7
8
Purpose
9
10
The GuideNet AI Control Framework provides a structured and implementable control architecture to govern Artificial Intelligence across its lifecycle.
11
It is designed to enable organisations to manage AI risk, demonstrate regulatory compliance, and support independent assurance.
12
13
14
Scope
15
16
The framework covers the end to end AI lifecycle including:
17
18
• Governance and oversight
19
• Data sourcing and management
20
• Model development and transparency
21
• Deployment, monitoring, and resilience
22
• Legal and regulatory compliance
23
• Third party and supply chain risk
24
• Evidence, auditability, and assurance
25
• Ethical and sustainability considerations
26
27
It applies to all AI systems, including predictive models, machine learning solutions, and generative AI.
28
29
30
31
Structure of the Framework
32
33
The framework is organised into:
34
35
• 12 control domains
36
• 88 defined control objectives
37
• Detailed operationalised controls
38
• Cross framework mappings
39
• Adoption stage applicability
40
41
Each control is uniquely identified and aligned to enterprise control expectations.
42
43
44
45
Framework Alignment
46
47
The framework is structurally aligned to leading standards and regulatory expectations including:
48
49
• NIST AI Risk Management Framework
50
• ISO 42001
51
• EU AI Act
52
• GDPR
53
• OECD AI Principles
54
55
Forward and reverse mapping enables traceability between regulatory obligations and control implementation.
56
57
58
59
Assurance and Audit Relevance
60
61
The framework is designed to support:
62
63
• Internal audit and independent assurance
64
• External audit and ICFR considerations
65
• Regulatory inspections and compliance reviews
66
• Risk management and governance oversight
67
68
Controls that impact financial reporting or audit reliance are identified through assurance flags.
69
70
71
72
Adoption Model
73
74
The framework is aligned to a four stage AI adoption model:
75
76
• Foundation
77
• Experimentation
78
• Operational
79
• Scaled
80
81
This enables organisations to implement controls progressively based on maturity, risk exposure, and reliance.
82
83
Refer to the “AI Adoption Model” tab for detailed definitions.
84
85
86
87
How to Use the RACM
88
89
The RACM tab provides the detailed control matrix and should be used to:
90
91
• Identify applicable controls based on AI use cases
92
• Map controls to regulatory and framework requirements
93
• Assess control design and implementation
94
• Support audit and assurance activities
95
96
Each row includes:
97
98
• Control ID and objective
99
• Risk addressed
100
• Operationalised controls