Polymorphic transformations occur in a minority of metals. Tungsten is not one of them. This observation does not rule out its intrinsic phase-transition possibility, but may be due to inaccessible thermodynamic and mechanical conditions in conventional experiments. Herein, we investigate the body-centered cubic (BCC) to face-centered cubic (FCC) transformation in tungsten via a rigorously validated machine learning potential (MLP) molecular dynamics (MD) approach. Our simulations demonstrate that strain-induced BCC to FCC transformation indeed occurs in tungsten. The FCC phase exceeds 20 at.% at low temperature and in the presence of dislocations. In the perfect crystal moderate temperatures favor phase transformation kinetics. Structural analysis confirms the transformation following the Bain and Nishiyama-Wassermann orientation relationship. These results establish the viability of strain-induced polymorphic transformation in tungsten and provide a critical guideline for the experimental production of duplex tungsten with enhanced ductility—addressing a longstanding challenge in tungsten material engineering for high-temperature structural applications.