Reinforcement Learning from Human Feedback
A training technique where human preferences are used as a reward signal to align AI models with human values and intended behaviour, commonly used to train large language models.
In plain language
Training AI by having humans rate its responses. The AI generates answers, humans say which are better and the AI learns from that feedback. It is like training a new employee through performance reviews and coaching.
Why this matters
RLHF is relevant to governance because it is the primary method used to align commercial AI systems with human values. Understanding how the AI models your organisation uses were trained helps you assess their risk profile and identify potential alignment gaps that governance controls must address.
Relevance
StrategyOrganisations using RLHF-trained models must understand the values embedded in their training process and assess whether those values align with the organisation's own risk tolerance, regulatory obligations and ethical commitments.
Related terms
Putting reinforcement learning from human feedback 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.