In April 2020, the IBM Japan branch of the Japan Metal Manufacturing, Information and Telecommunication Workers Union, which has employees at IBM Japan as members, filed a complaint to Labour Relations Commission claiming that IBM Japan refused to respond to collective bargaining in good faith by not disclosing information on AI-driven system for personnel evaluation and wage assessment.

According to the complaint, IBM Japan introduced Watson into wage assessments in 2019. Since then, the union repetitively asked IBM Japan to disclose information regarding the new system including AI's learning data and the output it displays to managers who assess it. IBM Japan has refused to do so by explaining the information is not intended to be disclosed and claiming that Watson is merely a support tool for managers.

A lawyer representing the union points out four possible adverse effects of the use of AI in this context, including the invasion of privacy, the increase of discrimination and unfair treatment, the black-boxing of personnel evaluation, and the likelihood of automation bias.

Business & Human Rights Resource Centre invited IBM Japan to respond to the alleged matter. The response is provided below.