AI at work -- Mitigating safety and discriminatory risk with technical standards