I cannot answer to the grammatical appropriateness of they (though others here seem to indicate it is becoming more acceptable). I can also agree with Dave Mulligan's answer that using the name ("Kris") in many situations would be most appropriate.
However, there are also other levels of appropriateness. I do not know how popular this opinion might be, but I value my freedom to classify a person as I want to and use the speech I want to refer to that person, no matter how they may want to be classified. That is, they have their right to classify themselves, but I have my own right to classify him or her differently (and classify him or her into a gender).
To me, this is entirely appropriate on other, non-grammatical levels (possibly philosophical, scientific, religious, etc., given circumstances), and if for no other reason that my perception.
And since it has long been grammatically appropriate to use "he" as a general or generic reference, (the Wikipedia article notes 18th century in textbooks, but examples in various languages go back thousands of years; see below), I may appropriately choose to do so still if I am seeking to make a generic reference (such would probably not be the case in the example given).
For example, Vern S. Poythress examined non-Biblical Greek texts (some "writings of Plutarch and Philo"), concluding:
Specific evidence nevertheless supports the idea that in Greek a male
“flavor” attaches to occurrences of third person masculine pronouns
and other masculine forms referring to human beings in a generic
statement. This evidence favors the conclusion that use of masculine
singular in English provides an appropriate match in meaning.
And early in the article had noted as well that:
Hebrew, Aramaic, and Greek show here [in some Biblical texts] a pattern similar to English: the
occurrence of the generic masculine sets up a perception that a male
example is being used to express a general principle
So grammatical precedent for the generic "he" is well founded, should one choose to continue to use it.
Whether one considers the arguments for not using it in a generic sense valid or not, and also in any specific instance for classifying, is really up to the one speaking/writing, not the one being referred to. It is the one communicating that is attempting to convey his or her thoughts about the matter, not the target referent's thoughts on the matter (unless that happens to be the goal of the one communicating).
And in the example, even being friends, one might disagree with the other's self-perception, and have to weigh whether it is appropriate or not to honor his or her perception over one's own.