GitHub published a postmortem for yesterday's outage already, and the cause surprised me: They blame scrapers, or rather ONE specific scraper, for taking down GitHub for 7 hours.

They may not need to go so far as this to protect their service, but I would accept that any access required occasional authentication. Unfettered, unlimited anonymous access seems dubious in this “I’m taking all your content” AI training world.