The lithium problem is Big Bang nucleosynthesis's one open wound. Standard BBN, using the baryon density Planck measures from the CMB, predicts the primordial abundances of the light elements, and for deuterium and helium-4 the predictions match observation beautifully. For lithium-7 they do not: the prediction runs a factor of 3 to 4 above the abundance measured in the atmospheres of the oldest metal-poor halo stars, where Li-7 sits on the famous Spite plateau, nearly constant across a wide range of metallicities, exactly as a primordial abundance should (Cyburt et al. 2008; Fields et al. 2014).
The plateau is what makes the problem stubborn. If stars destroyed their surface lithium, depletion should vary star by star with mass, rotation, and convection history, scattering the plateau rather than preserving it; uniform depletion mechanisms can be built but require tuning. Nuclear physics escape routes, revised reaction rates destroying Li-7's parent beryllium-7, have been progressively closed by laboratory measurement. And new-physics solutions during BBN must alter lithium while leaving deuterium and helium untouched, a surgical requirement few mechanisms meet. Four decades of effort have narrowed the options without selecting one.
The standing is unique in the tension catalog: a precision conflict between two of cosmology's most reliable calculations, with the leading suspicion now resting on stellar depletion physics, atomic diffusion and turbulent mixing slowly draining photospheric lithium, though no model yet earns the plateau's flatness without tuning.