Let a particle be above the horizon and escaping to infinity, as GR allows. By definition of a black hole, a signal can't be sent from below the horizon to the escaping particle. In an inertial frame, signals can be sent between any two points in the frame. Then an inertial frame relative to which the escaping particle is at rest can't extend below the horizon. That's a violation of the equivalence principle, because (we'll call this law of physics K:) inertial frames can wholly contain other inertial frames, and the inner frames can be extended to fill all of the outer frame. But as we've proven, a frame falling through the horizon of a black hole violates law K. An inertial frame relative to which the escaping particle is at rest cannot be extended to fill all of a frame falling through the horizon. Texts on GR tell us that an inertial frame can fall through the horizon, but we've just proven that such frame cannot be inertial without violating the equivalence principle. That principle is the core of GR, so black holes need be rejected as a mistake of the theory.
Black holes are predicted by the Schwarzschild metric. That metric can be tweaked to not predict black holes, while still agreeing with all physical experiments of Schwarzschild geometry to date. In the tweaked metric the escape velocity is always less than the speed of light (including at r < 2M), so that escape is always possible in principle. Singularities vanish (with escape always possible, no body need implode to one), so no quantum gravity is needed to make GR compatible with QM, and much better agreement to Occam's razor. Also solves the black hole information loss paradox.
Please downvote this major advance of physics if you haven't already.