Except Gemini, I suspect.
I’m sure that’s not gonna get misused. /s
I’m not a fan of LLMs, but an app has absolutely no business deciding what input method I use.
This feels too similar to completely idiotic practice of blocking the copy-paste of credentials and disallowing password managers in certain apps/websites.
It is impossible to really block those apps.
This is just for Google to look like they care.
Again you cannot know where text comes from.
Not sure how those AI apps work, but it would be a shame if this boils down to a list of google approved input apps, like gboard or Samsung keyboard and anything else gets blocked.
My guess, without reading the article, is that they will just require input apps on the Play Store targeting Android 16 with AI features to respect some sort of
allowAIInput = falseflagI’d wager side-loading and 3rd party app store apps will be ignored, but those apps will probably do it anyways to keep their Play Store releases.
I think that’s just a very small fix, intended so your AI assistant or voice input doesn’t pop up if it’s a password input field.
The documentation notes that “toolkits can optionally disable [AI writing tools] where not relevant e.g. passwords, number input, etc.”
seems correct
This is the only use case that makes sense to me. I can always use an AI tool and then type in the text, so this is certainly not a way to prevent AI from being used.