I think that many, though not all, committee members will try to do a good job, but they may fail. My perspective is mathematics in which a dissertation is very arcane (or can be, anyway) and few people in the world are as well versed in the particular tiny research area as the author, and, perhaps, the advisor. It may be a bit different in some other fields and perhaps wildly different in dissimilar fields.
But people are also busy with their own work and don't have the time or the inclination to learn every part of the background of a given (math) thesis. Many will try to read it, but will get stuck at some point. Questions from those people are likely to be to ask for some explanation of what is going on at the point that they got stuck. If you know your work well, these are fairly easy to answer, though it is possible to give an unsatisfactory, overly pedantic, answer. If you can provide insight about how to continue then you get a win.
But many people, being busy, will just defer to your advisor. This is especially the case if your advisor is a well respected mathematician. A couple of days (maybe hours) before the oral exam, they will go the the advisor and ask: "This is ok, right?". The advisor will assure them that all is well and so they have the confidence to sign off on your work. Proof by Authority, I suppose.
If you think this is cynical, then ask around. I know that it happens, both from being a student and from being an advisor. I won't admit, of course, to that behavior when I was another committee member.
However, as Bryan Krause says, things differ. You may get an eager committee member who has examined your work in detail. If they disagree with it or have a certain sort of personality, you might get uncomfortable questions. You need to give reasonable answers, of course, but you can, I hope also depend on your advisor to serve as your shield from overly critical things. You won't be required to give a perfect answer to every question, but need to demonstrate mastery of the ideas in the thesis and the surrounding micro-field.
But, in particular, you can't be expected to give snap answers to questions that require study and reflection to properly answer. If someone has seen a paper that you don't know about and it seems to say something different from what you have written, then it may take a lot of analysis to see (a) whether there really is a difference and (b) why the different results arose.
My committee was something like the following. My advisor knew everything that I did and he supported the work. Another faculty member was in the same research group and understood it well, but had few questions. But he had the background to follow it without a lot of detail. Another one, I think, just deferred to my advisor. Another who didn't ask questions in the "defense" came to me afterwards and admitted that he couldn't understand it very well (his sub field was different) but he found a few typos and gave me a list.
On the other hand, I heard of a case at the time, possibly apocryphal, in which a candidate in chemistry was denied a degree because the talk mentioned ph throughout. In that place, an external (out of field) examiner was required and that person asked for a lay-person's explanation of ph. The candidate froze up and couldn't answer, though it is a subject studied in first year chemistry. But, the fact that such a story was remembered (or made up) points out how rare such things really are.