AI GUIDE FOR BOARDS IN SINGAPORE
Section IV: Ethics and Trust
8.2.2 Enablers of a responsible AI culture
Board considerations
• Are responsible AI principles embedded into organisational values and day-to-day behaviours? • Do employees understand acceptable and unacceptable uses of AI? • Are incentives aligned with responsible adoption rather than uncontrolled experimentation? • Is there evidence that responsible AI principles are applied in practice? Prioritising speed without appropriate culture and governance can expose organisations to reputational, regulatory, and operational risks.
Responsible AI adoption depends on leadership, behaviours and organisational reinforcement mechanisms. Boards can support management through oversight of initiatives such as: • Training and awareness programmes. • Responsible AI policies and guidance. • Leadership role-modelling. • Clear escalation and reporting channels. • Incentive structures aligned with responsible outcomes. Innovation and AI experimentation should occur with appropriate organisational expectations and governance boundaries.
Responsible AI behaviours should be reinforced through leadership, incentives and organisational accountability.
8.3 Ecosystem Risks and Governance
Organisations building their AI capability increasingly rely on cloud providers, foundation model providers, so ware vendors, consultants, and specialist AI rms. While AI capabilities may be outsourced, accountability for outcomes remains with the organisation.
customer outcomes, employee decisions, operational processes, compliance obligations, or strategic decisions, organisations remain accountable for the consequences. Management should conduct AI-Speci c Due Diligence across six core areas: • Governance and accountability. • Cyber security and data protection. • Operational resilience. • Regulatory compliance. • Model transparency and explainability. • Vendor dependency risk.
8.3.1 ird-party risk
Where external AI systems inuence
77
Made with FlippingBook Digital Proposal Creator