Cooperation among network devices is envisioned to be a key enabler for the expected performance increase brought by the deployment of future 5G networks. In practice, new interesting opportunities for both smart signal processing and resource allocation in interference channels (ICs) are made possible by means of cooperative strategies. In this work, we propose a novel notion of cooperation for future 5G networks in which not only information but also energy is shared among the devices involved in the communication. In particular, we envision deployment policies for heterogeneous networks where transceivers may not have access to fixed power supplies. As a first study in this direction, we assess the impact of energy sharing paradigms by considering a target system of two multi-antenna transmitters serving their associated receivers by means of beamforming strategies based on a linear combination of maximal ratio transmission and zero forcing. We first characterize the achievable rate region for the resulting two-user multiple-input single-output (MISO) IC. Subsequently, relying on a parametric model of the system, we provide a closed-form expression for the optimal energy cooperation policy that maximizes the sum-rate with a QoS requirement on the transmitter that donates energy. Finally, numerical results show that parameters for which the energy cooperation policy preserves or improves the sum-rate of the two-user MISO IC can always be found.