NASICON-type Li1+xAlxTi2–x(PO4)3 (LATP), as a promising solid-state electrolyte for lithium-ion batteries, has received significant attention due to its simple preparation, low material cost, and good stability in water and air, but the formation of lithium dendrite greatly limits the applications. To elucidate the source of formation of lithium dendrite, in this study, the effects of Al content on the stability, electronic and Li+ mobility properties of the LATP surface with three Al doping concentrations (2AlTi, 4AlTi, 6AlTi) are investigated by combining first-principles calculations and molecular dynamics simulations. The LiTi2(PO4)3 (LTP) surface is also considered for comparison. The results indicate that the (012) surface terminated with Li atoms is the most stable facet. Further, the surface energy of LATP(012) decreases from 0.68 J/m2 to 0.43 J/m2 with the increase of Al content, suggesting that Al doping can effectively improve the stability of the LATP(012) surface. Electronic structure analysis reveals that the surface of LTP(012) retains the semiconductor properties consistent with the bulk phase, whereas the LATP(012) surface exhibits metallicity, which provides an electron pathway for forming the metallic Li . Consequently, the metallic characteristic of the LATP(012) surface is a reason for its lithium dendrite growth. For the Li+ transport properties, two different migration modes: vacancy migration and interstitial migration, are included. When Li+ migrates within the outermost surface, the migration barrier via vacancy is 1.67/1.69 eV for the LTP/LATP (012) surface, while the migration barrier via interstitial is 1.16 eV for LTP(012) and decreases from 1.31 to 0.87 eV with the increase of Al content for LATP(012). Obviously, within the outermost surface, Al doping can reduce the migration barrier of Li+. When Al doping concentration is 6AlTi, the migration barrier is lowest (0.87 eV). Nevertheless, the lowest migration barrier (0.87 eV) for Li+ on the LATP surface is significantly higher than its bulk minimum value of 0.34 eV. When Li+ migrates from the subsurface layer to the outermost surface, the migration barrier is 2.76 eV for LTP(012) and 2.05 eV, 3.20 eV, and 3.06 eV for LATP(012) with 2AlTi, 4AlTi, and 6AlTi content, respectively. All these migration barriers are greater than 2.00 eV, which prevents Li+ from migrating from the subsurface layer to the outermost surface for both LTP and LATP surfaces. Hence, the slow Li+ migration represents another important factor contributing to lithium dendrite growth on the LATP surface. Fortunately, increasing the Al doping concentration can reduce the migration barrier of Li+ and thus enhance its diffusion performance on the LATP surface. Molecular dynamics simulations further reveal that the diffusion behavior of Li+ on the LATP surface is influenced by a combination of factors, including Al content, Li+ occupancy, and ambient temperature. In particular, LATP(012)/6AlTi, LATP(012)/4AlTi, and LATP(012)/2AlTi possess their highest Li+ diffusion coefficients at 900 K, 1100 K, and 1300 K, respectively. Besides, Li+ near the Al doping site is easier to diffuse on the LATP(012) surface. Thus, our study indicates that by changing Al content, Li+ occupation positions, and the temperature, the Li+ diffusion performance of LATP(012) can be effectively modified, thereby suppressing the formation of lithium dendrites on the LATP(012) surface.