Interesting approach. We’ve seen similar spam/noise problems appear in financial workflow automation too — especially when AI-generated submissions scale faster than manual review processes.

too — especially

you can't trust org members either I have seen projects have inter maintainer fallouts. In general trust doesn't exist.

If companies can screw you over and claim it's a mistake, there isn't much a person can do.

It's all about level's of trust, a maintainer going rogue is less likely, a past contributor going rogue more likely but not too much, a stranger with a typo pr merged even more likely but still, a complete stranger least trust worthy.

You also can't fully trust org members because valid / "Trusted" accounts can be taken over by nefarious actors as well.

Actions runs from external contributors aren't run with Actions secrets; if you are using Actions right (i.e. not using pull_request_target wrong) you don't need to trust external contributors. (eta: iirc the original point of the Actions approval flow was preventing cryptomining spam from abusing free compute)