Outer Alignment
The problem of ensuring that the formal objective function or reward signal used to train an AI system accurately reflects the designer's true intentions and desired outcomes.
In plain language
Making sure the training goal you give an AI actually captures what you really want it to do. If you reward an AI for "maximising profit," it might learn to cut corners or cheat rather than build sustainable value.
Why this matters
Outer alignment failures can lead to systems that technically achieve their stated objective but produce harmful or undesirable outcomes. Organisations need to carefully define and validate their AI objectives to ensure they genuinely reflect business strategy and ethical principles.
Relevance
StrategyRequires clear definition of AI objectives that align with business strategy and values before development, preventing misalignment between stated goals and actual system behaviour.
Related terms
Putting outer alignment 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.