What if Artificial Intelligence will replace human in the Public Sector?

Every time I see the organizational and structural constraints in the public sector, I imagine the abovementioned scenario. There are many problems in the public area. Slow to adapt to changes, not being able to cut costs, obsolete services, and so forth. The consultancy firm suggested that the AI can help the public sector to reduce cost, increase speed, enhance reach, and focus more resources on frontline priorities. Of course, they will, albeit to the same extent that the other technologies brought about. But we know the public sector is also slow to adopt these innovations. Why?

Here is my hypothesis: when we use AI to address this question, the true value of AI will appear. The public sector tends to fail in improving their service delivery and innovating itself. What is the difference between the public sector and private sector, which often radically improve their effectiveness and decrease its costs? I think it is mainly because that person in the public sector is not meant to create new opportunities and the pi that directly return to them at the time of their evaluation period. Instead, their primary objective is to distribute resources to maximize social benefits of public people and eliminate inequity, not to maximize their return. In other words, their job is to eliminate the role of their job. This sounds good and very noble. But I feel that most of the people (probably except the saint-like Gundi and MLK) will inherently lose their motivation toward this noble purpose when they have to compromise their own benefits or job. Quite often, people unconsciously compromise the noble goal to create their opportunity and survive in the public sector. Unless you are truly saint or superhero, it is tough to eliminate your job to achieve the other goal. Therefore, in the public sector, any new technologies or innovative ideas are very hard to be adopted because there is a fundamental and structural lack of incentive to adopt it. How can we solve this problem to maximize social benefits in a radically effective and efficient manner? This will go back to my first assumption: What if Artificial Intelligence will replace human in the Public Sector?

The AI might be able to work on super hard until their roles cease to exist. If so, while we have to think how we should work when AI starts replacing all kinds of jobs (which always do), we also have to consider how we can replace civil servants with AI as fast and much as possible.