It feels like Google is increasingly becoming the interface layer above everyone else’s data.

For directory sites, I suspect simple lookup-style pages are most at risk of being absorbed into AI answers, while deeper/fresher/niche datasets may still have value if AI systems need reliable sources to cite or use.

On Google-Extended, I’m not convinced the separation between “Search” and “AI” will remain as clean as Google suggests now that Gemini is embedded everywhere across Search workflows.

Blocking may protect datasets short term, but there’s also a risk of becoming less visible in future AI-driven discovery layers.