What Is Frontier Model?
Frontier Model is a highly capable, general-purpose AI model at or near the leading edge of capability, whose risks are not yet fully understood.
Frontier Model — a highly capable, general-purpose AI model at or near the leading edge of capability, whose risks are not yet fully understood.
Frontier models are the largest, most capable general models — the ones whose novel capabilities (and potential for misuse) drive much of the current policy debate. Governments increasingly single them out for heightened scrutiny, including the work of national AI safety institutes, precisely because their behaviour is hard to predict in advance.
Source: Frontier AI policy (UK/US AI Safety Institutes)
Plain-language explanation
Frontier models are the largest, most capable general models — the ones whose novel capabilities (and potential for misuse) drive much of the current policy debate. Governments increasingly single them out for heightened scrutiny, including the work of national AI safety institutes, precisely because their behaviour is hard to predict in advance.
Related terms
See where you stand on AI governance
Take the free 7-question maturity assessment and get a personalised action plan.
Free assessment — 3 minutes →