Future quantum information networks will consist of quantum and classical agents, who have the ability to communicate in a variety of ways with trusted and untrusted parties and securely delegate computational tasks to untrusted large-scale quantum computing servers. Multipartite quantum entanglement is a fundamental resource for such a network and hence it is imperative to study the possibility of verifying a multipartite entanglement source in a way that is efficient and provides strong guarantees even in the presence of multiple dishonest parties. In this Letter, we show how an agent of a quantum network can perform a distributed verification of a source creating multipartite Greenberger-Horne-Zeilinger (GHZ) states with minimal resources, which is, nevertheless, resistant against any number of dishonest parties. Moreover, we provide a tight tradeoff between the level of security and the distance between the state produced by the source and the ideal GHZ state. Last, by adding the resource of a trusted common random source, we can further provide security guarantees for all honest parties in the quantum network simultaneously. Entanglement plays a key role in the study and development of quantum information theory. It has been widely used in all aspects of quantum information and has been essential to show the advantages obtained compared to the classical setting. Initially defined for bipartite states, the notion of entanglement has been generalized to multipartite systems and despite the complexity this notion acquires in this case, many interesting properties of multipartite entangled states are known. If we consider, for example, the quantum correlations of the Greenberger-Horne-Zeilinger (GHZ) state [1] and its nparty generalization, we can find a nonlocal game that can be won with probability 1 in the quantum setting, while any classical local theory can win the game with probability at most 3/4 [2].Multipartite entangled states are a fundamental resource when quantum networks are considered. Indeed, they allow network agents to create strong correlations in order to perform distributed tasks, to delegate computation to untrusted servers [3], or to compute, for example through the Measurement-Based Quantum Computation model [4]. A natural and fundamental question that arises then is whether the network agents should be required to trust the source that provides them with such multipartite entangled states or whether they are able to verify the entanglement.In this work, we show that a quantum agent can verify efficiently with respect to the necessary resources, that an untrusted source creates entanglement, even in the presence of dishonest parties.The model -We start our analysis by first describing in detail our model and its relation to previous work.Source: The source is untrusted. It is supposed to create the n-party GHZ state 1 √ 2 n |0 n + |1 n and distribute it to n parties. By applying a Hadamard and a phase shift ( √ Z) gate to each qubit, the GHZ state can be expressed by the locally equivale...