I think it’s a net positive as long as there’s a human in the loop. The key bit is this in my opinion:

While AI is playing a growing role in government work, officials say it is intended to assist, not replace, human workers — despite referring to such systems as “employees.” Futian’s regulatory framework requires each AI system to be monitored by a designated human supervisor to prevent errors and ensure compliance with ethical standards.

“The guardian of the AI-powered employee is responsible for overseeing its operation, and if any issues arise, the guardian is held accountable,” said Gao.

The AI isn’t the decision maker, it’s an automation tool that allows a human worker to do their job more efficiently, but responsibility still lies with the human.