ISO 42001: 2023 - A.5.5 Assessing Societal Impacts of AI Systems

ISO 42001 Control Description

The organisation shall assess and document the potential societal impacts of their AI systems throughout their life cycle.

Control Objective

To assess AI system impacts to individuals or groups of individuals, or both, and societies affected by the AI system throughout its life cycle.

Purpose

To identify and evaluate how AI systems may affect broader social structures, institutions, norms, and collective wellbeing beyond impacts on specific individuals or groups. Societal impact assessment enables organisations to understand systemic effects and contribute responsibly to social development.

Guidance on Implementation

What Constitutes Societal Impacts

Societal impacts are effects on:

• Social systems and structures
• Institutions and governance
• Cultural norms and values
• Collective behaviors and practices
• Environmental sustainability
• Democratic processes
• Economic systems
• Public trust and social cohesion

Types of Societal Impacts to Consider (ISO/IEC 42005 Clause 6.7, 6.8)

a) Democratic processes and civic engagement:

• Influence on electoral processes
• Impact on public discourse and information quality
• Effects on civic participation
• Surveillance and freedom of expression

b) Social cohesion and inequality:

• Widening or narrowing of social divides
• Impact on social mobility
• Reinforcement or disruption of power structures
• Distribution of benefits and harms across society

c) Environmental sustainability:

• Energy consumption and carbon footprint
• Resource depletion
• Electronic waste
• Climate change contribution or mitigation

d) Economic systems:

• Labor market transformation
• Industry disruption
• Wealth distribution
• Market concentration
• Economic growth or contraction

e) Cultural and social norms:

• Changes in social behaviors and practices
• Impacts on cultural traditions
• Shifts in values and norms
• Language and communication patterns

f) Public services and infrastructure:

• Access to healthcare, education, justice
• Quality and equity of public services
• Infrastructure resilience
• Service delivery effectiveness

g) Knowledge and information ecosystems:

• Information quality and reliability
• Spread of misinformation or disinformation
• Access to knowledge
• Media and journalism
• Scientific research practices

h) Public safety and security:

• Crime prevention and detection
• National security
• Public health
• Emergency response capabilities

i) Trust in institutions:

• Confidence in government, media, technology
• Legitimacy of decision-making processes
• Accountability mechanisms
• Transparency of systems

Considerations for Assessment (ISO/IEC 42001 Annex B.5.5)

The organisation should:

a) Consider scale and scope - Assess whether AI system deployment is:

• Limited or widespread
• Local, national, or global
• Temporary or long-term

b) Evaluate cumulative effects - Consider impacts when combined with other AI systems or technologies

c) Assess differential impacts - Understand how impacts vary across different societies, cultures, or regions

d) Consider intended and unintended consequences - Look beyond direct effects to secondary and tertiary impacts

e) Evaluate reversibility - Determine whether societal changes can be reversed if needed

f) Assess concentration of power - Consider whether the AI system concentrates decision-making, economic, or social power

Implementation Steps

1. Define societal scope - Identify which societies, communities, or social systems may be affected
2. Engage diverse perspectives - Consult sociologists, ethicists, policy experts, community representatives
3. Assess each impact category - Systematically evaluate potential societal impacts
4. Consider temporal dimensions:
   1. Short-term vs. long-term impacts
   2. Immediate vs. delayed effects
   3. Reversible vs. irreversible changes
5. Evaluate magnitude and likelihood - Determine significance of societal impacts
6. Identify mitigation and enhancement measures:
   1. Actions to prevent or minimise negative societal impacts
   2. Measures to maximise positive contributions to society
7. Document findings - Record societal impact assessment per Control A.5.3
8. Monitor societal effects - Establish ongoing monitoring for societal impacts post-deployment

Key Considerations

Complexity and uncertainty: Societal impacts are often difficult to predict with certainty. Use scenario planning, expert consultation, and historical analogies to inform assessment.

Long time horizons: Some societal impacts only manifest over years or decades. Consider longer-term effects beyond immediate deployment.

Systemic thinking: Society is a complex system. Small changes in AI systems can have nonlinear, cascading, or emergent effects.

Geographic and cultural variation: Societal impacts vary significantly across different contexts. An AI system may have positive impacts in one society and negative impacts in another.

Value judgments: Assessing whether societal impacts are positive or negative often involves value judgments. Be transparent about the values and frameworks guiding assessment.

Collective responsibility: While individual organisations conduct assessments, many societal impacts require collective action or policy responses. Engage with industry associations, policymakers, and civil society.

Precautionary approach: Where significant societal harms are possible but uncertain, consider precautionary measures.

Related Controls

Within ISO/IEC 42001:

• A.5.2 AI system impact assessment process
• A.5.3 Documentation of AI system impact assessments
• A.5.4 Assessing impact on individuals and groups
• Environmental considerations (system and computing resources - Control A.4.5)

Related Standards:

• ISO/IEC 42005:2025 Clause 6.7 (detailed guidance on societal impacts)
• ISO 26000 Social responsibility
• Environmental management standards (ISO 14001 series)