The algorithmic news has reconstructed the journalism and communication ecology, but also exposed some ethical problems, the core of which is the contradiction between "algorithmic black box" and the principle of news transparency, the contradiction between "algorithmic hegemony" and media justice, and the conflict between "algorithm manipulation" and delegated trust. To solve the ethical problems caused by algorithmic news, it is necessary to make clear that algorithmic news is a non-human "network actor", but not an independent ethical actor, and "human" is the main responsibility subject of algorithmic news. The logical starting point of algorithmic news governance should return to the benign interaction among technology, human and society. Algorithmic news governance should take "good" as the bottom line ethics and "up" as the responsibility for the free and comprehensive development of human beings and the orderly and healthy operation of society. Algorithmic news governance needs to establish a multi-agent collaborative governance system: the government should play a good role as a rule maker, process controller and bottom line controller; the platform institutions should bear the main responsibility of algorithmic news governance; the public should consciously avoid the blind or passive phenomenon in news consumption; the traditional media should strictly enforce the professional check and audit mechanism, and refuse the news production catering to human weakness and inertia. The common responsibility of the public and traditional media is to correct, improve and supervise the algorithm news.