I accidentally clicked a shady YouTube ad that mimics an iOS “Storage Full” alert, reported it (twice), and Google still said it was fine. Meanwhile, Gemini flags it in seconds for deceptive design...
This particular example of adversarial advertising by Google is directly targeting a flaw in a competitor’s (Apple) OS
The author asks why Google aren’t using their AI tools to combat these, which is a fair question. Surely they are using AI, of course, but no chance in hell are they using the same Gemini the public gets. It’s going to be tailored to their end goals, not the public-facing model the author screenshotted.
Two thoughts :