Frontier AI companies are hiring chemical and biological weapons experts to prevent destructive misuse of their platforms, but these mitigation systems have no external oversight — leaving critical risk governance subject to the whims of the companies and market forces, writes Javaid Iqbal Sofi.