Companies should really develop AI ethics frameworks which can be actionable, with apparent accountability and concrete ways of monitoring. Nonetheless, building code open source will not allow it to be comprehensible, which by many definitions ensures that the AI code is not transparent.