The transmission grid is a key enabler for the European low-carbon energy future. Large investments are required in the next years in order to strengthen the integration between national electricity systems and to replace ageing infrastructures, thus allowing the integration of a large share of fluctuating renewable electricity, mainly from wind and solar energy sources. In fact, renewable electricity – at least as far as wind power is concerned – is often produced in places away from main consumption centres and it needs to be transported over long distances; transmission is also needed to allow conventional generators cover the demand during low wind periods. R&D activities are more and more focused on the use of super-grids, off-shore grids, HVAC cables, HVDC technology and the perspective of using smart grids technologies. In this context, the adoption of multi-terminal (MT), possibly meshed, HVDC networks based on VSC (Voltage Source Converter) technology to convey large amounts of off-shore wind power into the bulk AC grid seems very appealing, even though several aspects are still under investigation (DC breaker technology, control paradigms, etc.). The introduction of Multi-Terminal Direct Current (MTDC) networks may determine new issues in power system operation, however, because large power injections are expected from VSC terminals into the AC grid. This would affect the security of the overall (AC and DC) system. To help operators manage these new scenarios characterised by significant variability and uncertainty, new analysis techniques based on risk concept have been introduced in recent years. This paper presents the application of a probabilistic risk-based approach to analyze the effect of power injections from MT-HVDC networks on the security of the AC bulk system, evaluated in terms of risk of loss of load, high currents and low voltages. In particular, risk assessment is performed for different load levels in the grid, and different amounts and configurations of MT-HVDC power injections, in order to evaluate the dependence of the proposed risk indices on the mentioned quantities which are important for AC system operation.