Enough of these "falsified science" stories come out and people will become interested in stories debunking bad science. I subscribe to a podcast called "everything hertz" for this reason.

https://everythinghertz.com/

https://twitter.com/jamesheathers

https://twitter.com/dsquintana

By convincing the people deciding about their career advancement that negative results are worth something too. (no, that isn't easy, since it's the same people to a good degree, which is a problem with solving many problems in academia)

Give groups larger grants for publishing negative results too? More money seems to be workable enough in the rest of the economy. Obviously it would be best if those at the top make the change and let it cascade down. But any private source of funds could also experiment with changing the grant model.