Together with inactivated influenza vaccines (IIV), live attenuated influenza vaccines (LAIV) are an important tool to prevent influenza A virus (IAV) illnesses in patients. LAIVs present the advantages to have a needle-free administration and to trigger a mucosal immune response. LAIV is approved for healthy 2- to 49-year old individuals. However, due to its replicative nature and higher rate of adverse events at-risk populations are excluded from the benefits of this vaccine. Using targeted mutagenesis, we modified the nonstructural protein 1 of the currently licensed LAIV in order to impair its ability to bind the host cellular protein CPSF30 and thus its ability to inhibit host mRNA poly-adenylation. We characterized our optimized LAIV (optiLAIV) in three different mouse models mimicking healthy and high-risk patients. Using a neonatal mouse model, we show faster clearance of our optimized vaccine compared to the licensed LAIV. Despite lower replication, optiLAIV equally protected mice against homosubtypic and hetesubtypic influenza strain challenges. We confirmed the safer profile of optiLAIV in Stat1^-/-mice (highly susceptible to viral infections) by showing no signs of morbidity compared to a 50% mortality rate observed following LAIV inoculation. Using a human nasal 3D tissue model, we showed an increased induction of ER stress-related genes following immunization with optiLAIV. Induction of ER stress was previously shown to improve antigen-specific immune responses and is proposed as the mechanism of action of the licensed adjuvant AS03. This study characterizes a safer LAIV candidate in two mouse models mimicking infants and severely immunocompromised patients and proposes a simple attenuation strategy that could broaden LAIV application and reduce influenza burden in high-risk populations.IMPORTANCELive attenuated influenza vaccine (LAIV) is a needle-free, mucosal vaccine approved for healthy 2- to 49-year old individuals. Its replicative nature and higher rate of adverse events excludes at-risk populations. We propose a strategy to improve LAIV safety and explore the possibility to expand its applications in children under 2-year old and immunocompromised patients. Using a neonatal mouse model, we show faster clearance of our optimized vaccine (optiLAIV) compared to the licensed LAIV. Despite lower replication, optiLAIV equally protected mice against influenza virus challenges. We confirmed the safer profile of optiLAIV in Stat1^-/-mice (highly susceptible to viral infections) by showing no signs of morbidity compared to a 50% mortality rate from LAIV. OptiLAIV could expand the applications of the current LAIV and help mitigate the burden of IAV in susceptible populations.