Abstract: Since ancient times, insurance has been playing an important societal role of hedging away uncertainty associated with "insurable risk". Without it many socio-economic activities would be deemed too risky and impossible to undertake. Those receiving insurance cover (insureds) rely on insurer's ability to honour insurance claims when they occur. On the other hand, whilst satisfying minimum solvency requirements imposed by regulators over a one year period to protect policyholders and maintain stability of the insurance market, insurers also have natural incentives to strategise their risk taking and apply active risk management over a longer time horizon to ensure their business is sustainable and such that adds value to investors (shareholders) that provide vital paid-in capital. These additional risk management incentives are mainly due to the following differences between insurance entities and conventional financial organisations: 1. Unlike typical financial organisations, insurers leverage themselves via issuing "risky debt" in the form of insurance policies - here additional riskiness is associated with uncertainty around occurrence time and severity of insurance event; and 2. Insurance risks are more skewed towards downside and heavier in the tail when compared to financial risks. This talk covers the following selected topics of Enterprise Risk Management (ERM) in insurance and discusses how insurers could efficiently use their ERM tools, internal models and related processes to navigate towards the optimal use of capital resources and enhanced shareholders value: A. Identifying entity's "Enterprise Risk" and defining "Risk Appetite" and related risk levers / controls of capital optimisation. B. Leapfrogging the traditional "CoV paradigm" of measuring riskiness of insurance results - use of alternative efficient risk measures to excessively skewed / heavy-tailed insurance risk profiles. C. Multi-period dynamic optimisation of insurance risk taking and capital resources - use of numerical techniques of dynamic programming along with computational power of computers. D. Managing model risk or "knowing your unknowns" (Knightian uncertainty) - use of quantitative methods of robust decision control under ambiguity averseness.