Mudrock samples were investigated from two fault zones at ∼3066 m and ∼3296 m measured depth (MD) located outside and within the main damage zone of the San Andreas Fault Observatory at Depth (SAFOD) drillhole at Parkfield, California. All studied fault rocks show features typical of those reported across creep zones with variably spaced and interconnected networks of polished displacement surfaces coated by abundant polished films and occasional striations. Electron microscopy and X-ray diffraction study of the surfaces reveal the occurrence of neocrystallized thin film clay coatings containing illite-smectite (I-S) and chlorite-smectite (C-S) minerals. 40Ar/39Ar dating of the illitic mix-layered coatings demonstrated Miocene to Pliocene crystallization and revealed an older fault strand (8 ± 1.3 Ma) at 3066 m MD, and a probably younger fault strand (4 ± 4.9 Ma) at 3296 m MD. Today, the younger strand is the site of active creep behavior, reflecting a possible (re)activation of these clay-weakened zones. We propose that the majority of slow fault creep is controlled by the high density of thin (<100 nm thick) nanocoatings on fracture surfaces, which are sufficiently smectite-rich and interconnected at low angles to accommodate slip with minimal breakage of stronger matrix clasts. Displacements occur by frictional slip along particle surfaces and hydrated smectitic phases, in combination with intracrystalline deformation of the clay lattice, associated with extensive mineral dissolution, mass transfer, and residual precipitation of expandable layers. The localized concentration of smectite in both I-S and C-S minerals contributes to fault weakening, with fracturing and fluid infiltration creating new nucleation sites for neomineralization on displacement surfaces during continued faulting. The role of newly grown, ultrathin, hydrous clay coatings contrasts with previously proposed scenarios of reworked talc and/or serpentine phases as an explanation for weak fault and creep behavior at these depths.