There is currently no efficient process for going from the oxide to the metal. The only processes used at scale involve reducing titanium chloride with more reactive metals (sodium or magnesium). They are batch processes and there are significant waste issues and inefficiencies having to consume a more reactive metal to make the metallic titanium.

Reducing aluminum chloride with sodium is how aluminum was produced before direct electrolysis, and it too was a much more expensive process: https://en.wikipedia.org/wiki/History_of_aluminium#Early_ind...

A company named Metalysis has been working on an electrolytic process starting with oxide for a long time: https://www.enterprise.cam.ac.uk/news/metalysis-successfully...

The articles are dead or just impossible to find now, but at the time I recall reading about a vicious intellectual property fight between the Metalysis founders and some of the other researchers who had worked on it at Cambridge. My take was that Metalysis won the IP fight but the people who lost had better hands-on knowledge of how to make the process work, so everyone lost in the end. Metalysis took forever to actually offer titanium powder as a product even though it was "almost ready" for a decade.

The other issue affecting Metalysis product, which might have persisted even if not for the IP fight, is that it has more trace oxygen in it than titanium made in the traditional batch processes. This is important because titanium loses ductility at even low levels of oxygen incorporation.