For the short term and for practical purposes, the laws of physics do not change. But the laws of physics represent only our current perception of reality. The laws change when our understanding of them changes. Once we declared the world flat and that was reality. Now we say it is spherical and we have the evidence to prove that it is. Some time later, we may decide to change that perception based on new data.

I do not think we can give a provable answer to this. If the cosmos was created, the Creator could probably alter the rules at will. If the cosmos is one possible configuration of proto-matter/energy it is conceivable that this form could morph into some other form. If string theory is correct, there may be some interaction with other strings or branes that could alter physical behaviors in ways we have not seen or imagined.

In so much as our understanding of them alters yes, in terms of actual "laws" of nature probably not. However there is some research into whether things such as fundamental constants (such as the speed of light, charge of an electron etc) may change over billions of years. There is a group at Cambridge working on it now I believe.

Well the "laws" are only abstractions of the data we've collected by observation. They very well may be, and in some cases have turned out to be, mostly true or simplifications. The more we know about physics, the more we learn there are no hard and fast laws, and we also learn that we should be cautious about putting a "period" at the end of a statement of physics "truth". In the late 1800s, we really thought we were almost at a complete understanding of the universe - there were just a few pesky unexplained phenomena that we would certainly figure out soon, the photoelectric effect (that led to quantum mechanics and its weirdness, and the death of classical physics with its pure, non-probabilistic determinism) and the fact that the speed of light seemed to need to be represented as a constant, despite Newtonian mechanics suggesting that it, like any other velocity, must be subject to revision when faced with motion relative to light. Einstein and other physicists demonstrated light has wave and particle like aspects, depending on the measurement, and that the speed of light is constant regardless of one's state of motion, which means time "flows" at a slower rate and lengths contract for the faster moving object, when compared to an observer closer to rest with respect to that observer. And, as weird as it all seems, it turned out both of these predictions can be repeatedly experimentally verified. The universe is weird, and evolution did not prepare us for how much. As we learn more, the "laws" will change. We don't even know if the laws are constant or in flux - we never can know, since we always have to go under the assumption that there is an underlying constant order that can be expressed as a theory that doesn't change, and we'll take all the crazy observations and experimental evidence and try to come up with an even crazier theory that makes it all come together in the most elegant way possible. Even armed with a so-called "theory of everything", we may find, even if it takes a million years of this theory being perfectly accurate at high levels of precision at explaining and predicting all observed phenomena, that someday something is observed that throws the whole theory into question. More frustratingly, we may find competing theories that are equally elegant and equally successful, and there may ultimately be no way to resolve this embarassment of riches. We're not there yet though. We may yet find new laws of physics - one idea gaining ground is a law of "scale relativity", that the laws of physics are not only the same for any observer regardless of state of motion, but also regardless of scale / size. Some people mean, by "laws" of physics, the constants. Well, there is a lot of evidence that some of the constants, such as the speed of light, may fluctuate over long periods of time. The fine structure constant, alpha, which depends on the speed of light, has changed by a small amount over the course of the history of the universe. Although it deeply complicates our theories, some physicists have responded by representing one or more of the so-called fundamental constants in their equations as vector-fields, so that theories allow for them to change over time while retaining their structure and accuracy. It is unclear how much of the constants are really variable but barely noticeable as being variable because the variability is so small, how many are really constant, if any, and how we might explain and represent such variability. My gut instinct is that we'll find that every constant turns out to be only a close approximation of a variable vector-field.