Oooof, looks like the primary way of creating it is through... fission reactors.

https://en.wikipedia.org/wiki/Tritium#Production

And there's not much of it:

> According to a 1996 report from Institute for Energy and Environmental Research on the US Department of Energy, only 225 kg (496 lb) of tritium had been produced in the United States from 1955 to 1996.[a] Since it continually decays into helium-3, the total amount remaining was about 75 kg (165 lb) at the time of the report.

The concentration of deuterium in the ocean is about 150-160 parts per million and with 1233.91 quintillion liters covering the earth we have approximately 8.2260667e+12kg worth of it to extract, so we've got a bit to work through!

Tritium however is far more rare with only trace amounts of it being available within nature and barely more than a kg produced per year. Producing the 100s of kgs required per year still seems to be an unsolved problem, although my quick searching shows there's a couple viable solutions for it.

The solution is that fusion power plants can breed tritium and become net producers of it...

Though in practice enough will be lost that probably they'll still be somewhat net consumers-- just not nearly to the extent predicted by a simple thermodynamic model.

Still, even if fusion becomes a net producer of tritium, the whole tritium-is-hard-to-get problem will likely be a constraint that we'll be fighting as we ramp up use of fusion power in the future.