AI Safety
The discipline of ensuring AI systems perform their intended function without causing unintended harm, encompassing alignment, robustness and fail-safe design.
In plain language
Ensuring AI does what it's meant to do and nothing dangerous. For example, a self-driving car that doesn't swerve into oncoming traffic, or a medical chatbot that refuses to give treatment advice. It's about preventing foreseeable failures.
Why this matters
AI safety is a strategic imperative that extends beyond technical implementation. Boards and executives must understand safety risks, ensure governance frameworks include safety requirements and testing protocols, and establish incident response procedures that reflect the severity of potential harms.
Relevance
GovernanceSafety requirements must be embedded in governance frameworks, project approval criteria and operational controls to ensure organisations can demonstrate accountability for AI system performance.
Related terms
Putting ai safety into practice in your organisation?
Ready to transform your AI strategy?
Partner with Australia's AI strategy and governance specialists. From adoption roadmaps to ISO 42001 audit readiness.