I'm actually growing increasingly suspicious that you personally haven't actually tested OP's claims.. Have you? Or am I literally the only one in this thread who tested anything lol

There's literally no actual evidence what they're saying is true, and you're making assumptions that its an A/B test. We don't even know what projects they tested (for all we know, they tested a project where the wiki was restricted to the team only, and assumed it affected everyone)

Either that or OP is just simply wrong, or, was affected by the outage that conveniently happened the day OP posted this, which specifically mentioned things including pages.. You can see outages on https://www.githubstatus.com/history . You're assuming again it shouldn't affect other things..

They didn't even post any information on what repos they tested. We see these crazy witch-hunts constantly in the tech community. Remember the Xbox ring of death debacle where people told others that Microsoft was stupid and left a piece of paper in their heatsinks? Turned out it was a thermal pad.

What data do you think Microsoft gains from "datamining" searches in repos?

Also, there are huge open source projects on Github, and if the searching thing was a big deal at all, they would be making public announcements.. They aren't. And again, the people making a deal out of it I've looked up, haven't contributed much either (so, it feels like they simply are using it to attack Microsoft).

I just tested 3 other repos and they all have wiki's and such working publicly. Given op is the only one I've seen complain about this anywhere and hasn't posted any evidence, I think its just a weird witch hunt.. Either that, or it could even be done in a specific country for legal reasons too.. But there is no way of even testing that because we don't know where OP is (they didn't say that). Or it could be done to reduce server load

I sure as hell don't agree with Copilot scraping repos, but there is no actual evidence in this discussion thread, only a claim by OP