You can determine/define a minimum group size, and regulation that prohibits finer-grained targeting.

This is something that happens in many other fields, for example to prevent insurance policies from becoming individualized cost-spreading (with front-loading) plus profit-margin.

(Or, should be happening. And this is one of the battles around regulation. When people see their costs skyrocketing, they should ask themselves whether that is the "cost of regulation", or rather the cost of its absence.)

It can become quite difficult to micromanage through regulation. Sometimes, that's necessary, e.g. with pollutants whose individual effects vary dramatically.

But sometimes, it's possible to observe boundaries beyond which negative effects become quite pronounced with escalating proclivity. You draw the line there, or a bit on the safe side, and say, "Thou shall not pass."

If people get the choice, as opposed to being dictated to, I think that they are going to find that in most domains, they don't want to be micro-targeted. And maybe, we will come up with laws and regulations to stop it -- or substantially hinder it, at least.

It can even feed into a healthy society. How will you ever learn of, and possibly experience, anything new and "out of your comfort zone", if you are constantly being algorithmically channeled back into it?