This is pretty easy to solve. If you present data by algorithm, you are no longer an impartial common carrier and are liable for the content you present. If the user decides you don’t, ala social media 1.0.
This is one of those things that don’t translate to legal reality very well, as then you have to define “what is an algorithm”. Is adding advertisements an algorithm? Is including likes an algorithm? Is automatically starting the next video after a previous one has finished an algorithm? Is infinite scroll an algorithm? Etc
The said regulation can then mandate that after calibration and correction the feed pull back by training the algorithm to adjust it in a rapid A/B test.
This is all doable by the companies themselves, but since they wont, the key is to mandate it and publish the aggregate results regularly — like make it part of the quarterly share holder's SEC reporting requirement or something.