The possibility and potential impact of a nuclear accident are a subject of debate and a key factor in public concern for nuclear facilities virtually since the first reactor. Although nuclear power plants have several built-in physical barriers to maintain the safety of the system, they are designed to prevent radioactive isotopes from escaping into the environment. However, the experiences for past few decades indicate that nuclear accidents can happen. The radioactive waste produced by nuclear power plants also threatens the natural environment and human health. It can exist in the natural environment for up to 100 000 years. Deep geological disposal of spent fuel and radioactive waste is the most feasible and safe option. The permanent disposal of highly radioactive nuclear waste is to place it in a container that can be isolated from the natural environment until the radioactivity of the fission products reducing to a safe level. The short-term and long-term corrosion mechanism of glass is one of the key points. The current research is to use various methods to simulate the short-term and long-term behaviors of the glass body in different geological environments. However, there is a lack of effective connections between various experiments. In just a few decades after the vitrified body is buried, countless problems that are not considered in the disposal of waste glass are discovered. The current situation is that existing research results lag far behind actual needs. From this point of view, research related to the corrosion of solidified glass is not a precautionary measure, but an urgent one that cannot be delayed.Borosilicate glass is currently a material of choice internationally for immobilizing high-level nuclear waste, including excess plutonium from dismantled nuclear weapons and highly radioactive liquid/solid waste from spent fuel reprocessing. In geological repositories, nuclear waste containing borosilicate glass is embedded in a multi-barrier containment system that should prevent water from entering the glass and the release of radioactive materials. However, groundwater corrosion of glass over long-term storage beyond geological time scales cannot be ruled out, so experimental glass corrosion studies, especially those dealing with corrosion mechanisms, are crucial to assess the long-term performance of glass. After long-term observation and experiments, the corrosion of glass in an aqueous solution environment is often divided into five stages, i.e., interdiffusion between solid and liquid stage, rapid initial reaction with the contact solution stage, corrosion rate decreases stage, residual rate stage and alteration recovery stage. Although these behaviors are not experienced by all materials under all corrosion conditions, one or more of the above can perfectly describe the corrosion process of the material. Three reactions usually occur during the reaction process, i.e., (1) hydration reaction (diffusion), water enters the glass as a complete molecule, (2) hydrolysis reaction, water reacts with the metal-oxygen bond in the glass to generate hydroxyl groups, and (3) ion exchange reaction, the modified cations in the glass are replaced by protons or other cations in the water. There is still controversy for the mechanism that controls the decrease in corrosion rate, but it is basically divided into two situations. One is due to the increase in the concentration of silicon at the interface, which causes the leaching rate to decrease, and another is due to the formation of a corrosion layer at the interface. Large molecules cannot be transferred. There are currently three most competitive models that are widely accepted. The first model is a chemical affinity model, which is based on the concept of "deviation from equilibrium" and can solve the problem of dissolution of the glass network, changes in pH value, precipitation of stable or metastable reaction products, and secondary problems at the interface (reaction zone) between glass and solution. Issues such as silicon oxide saturation and residual affinity of long-term reactions under near-saturated conditions; the second model is an alkali-proton exchange model to form a corrosion layer based on ion exchange reactions. The corrosion layer is considered as the ion exchange of protons with network modifiers such as Na+. The reaction produces products with silanol (Si—OH) groups. The silanol group is repolymerized to form the residual hydrated glass layer. The third model is a gel layer model, which believes that the corrosion layer formed can effectively protect the remaining glass from being corroded. It proposes the release of substances in the glass into the solution, that is, the corrosion rate is determined by the transmission characteristics of the "gel layer".Summary and prospects The corrosion of nuclear waste glass over long-time is undoubtedly the most complex problem in glass science to date. Because its components include dozens of oxides, there are cracks formed during the cooling stage, the geometry of the glass block is complex, the thermal, chemical and water boundary conditions are time-varying, involving many coupling phenomena. In addition, the time scale of deep geological disposal security assessment goes beyond direct verification. There is therefore a need to develop one or more rigorous, multi-scale theories to integrate basic understandings into models, which can be verified through dedicated experiments and compared with corrosion in natural or archaeological glass. Future models predicting a long-term corrosion of silicate glasses in aqueous environments should consider the spatial and temporal coupling of glass dissolution and silica precipitation and growth, controlling the transport of chemicals through the growing corrosion zone. Some specific issues include the evolution of porosity and structure over time in SAL, and how to control the transfer of materials between the aqueous solution and the glass; in the actual process, the nuclear fission products and actinides in the nuclear waste glass are what is the reaction during the glass corrosion process. Radioactive elements generate a large amount of heat during the placement process, and the ion irradiation they generate cause damage to the glass structure and SAL. The nucleation and growth of silicon dioxide at the reaction interface affect the glass dissolution rates and the use of ceramics and glass-ceramics for deep geological disposal of radioactive nuclear waste.