On "Meta Drops Training From Its Crawl Clause And Keeps The Crawl"

Meta's move is smart PR cover for what amounts to a non-concession. They removed "training" from the language while keeping the right to crawl—which means they still get the data, just with cleaner optics when defending the practice. For practitioners, this changes nothing about the actual risk: if your site is accessible to Meta's crawlers and you're a business user, assume your content is being used for model improvement regardless of what the terms say. The semantic shift buys them deniability without sacrificing the asset.