XPath and XSLT are still active, especially in the publishing space where documents are not stored in HTML, but in other formats like JATS or DocBook and need converting to HTML when displaying to the user, or converting to other XML formats when interfacing with other formats/vendors like crossref. It's also still used for things like processing the US bill XML data (e.g. https://www.govinfo.gov/bulkdata/BILLS/resources).

Does anyone still use XML? I would imagine the answer to both is yes. I often wish JSON had something as good as xpath.